publication_date
stringlengths 20
20
| doc_url
stringlengths 33
34
| id
stringlengths 33
34
| authors
stringlengths 5
20.1k
| update_date
stringlengths 20
20
| category_all
stringlengths 5
138
| abstract
stringlengths 123
3.09k
| category_primary
stringclasses 78
values | title
stringlengths 10
211
|
|---|---|---|---|---|---|---|---|---|
2024-05-30T01:11:35Z | http://arxiv.org/pdf/2405.19592v1 | http://arxiv.org/abs/2405.19592v1 | Zhenmei Shi, Junyi Wei, Zhuoyan Xu, Yingyu Liang | 2024-05-30T01:11:35Z | cs.LG, cs.AI, cs.CL | Large language models (LLM) have emerged as a powerful tool for AI, with the
key ability of in-context learning (ICL), where they can perform well on unseen
tasks based on a brief series of task examples without necessitating any
adjustments to the model parameters. One recent interesting mysterious
observation is that models of different scales may have different ICL
behaviors: larger models tend to be more sensitive to noise in the test
context. This work studies this observation theoretically aiming to improve the
understanding of LLM and ICL. We analyze two stylized settings: (1) linear
regression with one-layer single-head linear transformers and (2) parity
classification with two-layer multiple attention heads transformers (non-linear
data and non-linear model). In both settings, we give closed-form optimal
solutions and find that smaller models emphasize important hidden features
while larger ones cover more hidden features; thus, smaller models are more
robust to noise while larger ones are more easily distracted, leading to
different ICL behaviors. This sheds light on where transformers pay attention
to and how that affects ICL. Preliminary experimental results on large base and
chat models provide positive support for our analysis. | cs.LG | Why Larger Language Models Do In-context Learning Differently? |
2024-05-29T23:45:42Z | http://arxiv.org/pdf/2405.19575v1 | http://arxiv.org/abs/2405.19575v1 | Umar Ibrahim, Abubakar Yakubu Zandam, Fatima Muhammad Adam, Aminu Musa | 2024-05-29T23:45:42Z | cs.CL, cs.AI | Aspect-based Sentiment Analysis (ABSA) is crucial for understanding sentiment
nuances in text, especially across diverse languages and cultures. This paper
introduces a novel Deep Convolutional Neural Network (CNN)-based model tailored
for aspect and polarity classification in Hausa movie reviews, an
underrepresented language in sentiment analysis research. A comprehensive Hausa
ABSA dataset is created, filling a significant gap in resource availability.
The dataset, preprocessed using sci-kit-learn for TF-IDF transformation,
includes manually annotated aspect-level feature ontology words and sentiment
polarity assignments. The proposed model combines CNNs with attention
mechanisms for aspect-word prediction, leveraging contextual information and
sentiment polarities. With 91% accuracy on aspect term extraction and 92% on
sentiment polarity classification, the model outperforms traditional machine
models, offering insights into specific aspects and sentiments. This study
advances ABSA research, particularly in underrepresented languages, with
implications for cross-cultural linguistic research. | cs.CL | A Deep Convolutional Neural Network-based Model for Aspect and Polarity Classification in Hausa Movie Reviews |
2024-05-29T23:19:28Z | http://arxiv.org/pdf/2405.19567v1 | http://arxiv.org/abs/2405.19567v1 | Shenghuan Sun, Gregory M. Goldgof, Alexander Schubert, Zhiqing Sun, Thomas Hartvigsen, Atul J. Butte, Ahmed Alaa | 2024-05-29T23:19:28Z | cs.AI, cs.CL, cs.CV, cs.LG | Vision-Language Models (VLM) can support clinicians by analyzing medical
images and engaging in natural language interactions to assist in diagnostic
and treatment tasks. However, VLMs often exhibit "hallucinogenic" behavior,
generating textual outputs not grounded in contextual multimodal information.
This challenge is particularly pronounced in the medical domain, where we do
not only require VLM outputs to be accurate in single interactions but also to
be consistent with clinical reasoning and diagnostic pathways throughout
multi-turn conversations. For this purpose, we propose a new alignment
algorithm that uses symbolic representations of clinical reasoning to ground
VLMs in medical knowledge. These representations are utilized to (i) generate
GPT-4-guided visual instruction tuning data at scale, simulating clinician-VLM
conversations with demonstrations of clinical reasoning, and (ii) create an
automatic reward function that evaluates the clinical validity of VLM
generations throughout clinician-VLM interactions. Our algorithm eliminates the
need for human involvement in training data generation or reward model
construction, reducing costs compared to standard reinforcement learning with
human feedback (RLHF). We apply our alignment algorithm to develop Dr-LLaVA, a
conversational VLM finetuned for analyzing bone marrow pathology slides,
demonstrating strong performance in multi-turn medical conversations. | cs.AI | Dr-LLaVA: Visual Instruction Tuning with Symbolic Clinical Grounding |
2024-05-29T23:11:53Z | http://arxiv.org/pdf/2405.19563v1 | http://arxiv.org/abs/2405.19563v1 | Michael Fore, Simranjit Singh, Chaehong Lee, Amritanshu Pandey, Antonios Anastasopoulos, Dimitrios Stamoulis | 2024-05-29T23:11:53Z | cs.CL | Misinformation regarding climate change is a key roadblock in addressing one
of the most serious threats to humanity. This paper investigates factual
accuracy in large language models (LLMs) regarding climate information. Using
true/false labeled Q&A data for fine-tuning and evaluating LLMs on
climate-related claims, we compare open-source models, assessing their ability
to generate truthful responses to climate change questions. We investigate the
detectability of models intentionally poisoned with false climate information,
finding that such poisoning may not affect the accuracy of a model's responses
in other domains. Furthermore, we compare the effectiveness of unlearning
algorithms, fine-tuning, and Retrieval-Augmented Generation (RAG) for factually
grounding LLMs on climate change topics. Our evaluation reveals that unlearning
algorithms can be effective for nuanced conceptual claims, despite previous
findings suggesting their inefficacy in privacy contexts. These insights aim to
guide the development of more factually reliable LLMs and highlight the need
for additional work to secure LLMs against misinformation attacks. | cs.CL | Unlearning Climate Misinformation in Large Language Models |
2024-05-29T23:08:31Z | http://arxiv.org/pdf/2405.19562v1 | http://arxiv.org/abs/2405.19562v1 | Lucas Monteiro Paes, Dennis Wei, Flavio P. Calmon | 2024-05-29T23:08:31Z | cs.CY, cs.CL, cs.LG | Feature attribution methods explain black-box machine learning (ML) models by
assigning importance scores to input features. These methods can be
computationally expensive for large ML models. To address this challenge, there
has been increasing efforts to develop amortized explainers, where a machine
learning model is trained to predict feature attribution scores with only one
inference. Despite their efficiency, amortized explainers can produce
inaccurate predictions and misleading explanations. In this paper, we propose
selective explanations, a novel feature attribution method that (i) detects
when amortized explainers generate low-quality explanations and (ii) improves
these explanations using a technique called explanations with initial guess.
Our selective explanation method allows practitioners to specify the fraction
of samples that receive explanations with initial guess, offering a principled
way to bridge the gap between amortized explainers and their high-quality
counterparts. | cs.CY | Selective Explanations |
2024-05-29T23:07:58Z | http://arxiv.org/pdf/2406.00062v1 | http://arxiv.org/abs/2406.00062v1 | David Pissarra, Isabel Curioso, João Alveira, Duarte Pereira, Bruno Ribeiro, Tomás Souper, Vasco Gomes, André V. Carreiro, Vitor Rolla | 2024-05-29T23:07:58Z | cs.CL, cs.AI, cs.CR, cs.LG, I.2.7 | Automated clinical text anonymization has the potential to unlock the
widespread sharing of textual health data for secondary usage while assuring
patient privacy and safety. Despite the proposal of many complex and
theoretically successful anonymization solutions in literature, these
techniques remain flawed. As such, clinical institutions are still reluctant to
apply them for open access to their data. Recent advances in developing Large
Language Models (LLMs) pose a promising opportunity to further the field, given
their capability to perform various tasks. This paper proposes six new
evaluation metrics tailored to the challenges of generative anonymization with
LLMs. Moreover, we present a comparative study of LLM-based methods, testing
them against two baseline techniques. Our results establish LLM-based models as
a reliable alternative to common approaches, paving the way toward trustworthy
anonymization of clinical text. | cs.CL | Unlocking the Potential of Large Language Models for Clinical Text Anonymization: A Comparative Study |
2024-05-29T23:06:54Z | http://arxiv.org/pdf/2405.19561v1 | http://arxiv.org/abs/2405.19561v1 | Venkat Venkatasubramanian, Arijit Chakraborty | 2024-05-29T23:06:54Z | cs.AI, cs.CL, I.2.0; I.2.7 | The startling success of ChatGPT and other large language models (LLMs) using
transformer-based generative neural network architecture in applications such
as natural language processing and image synthesis has many researchers excited
about potential opportunities in process systems engineering (PSE). The almost
human-like performance of LLMs in these areas is indeed very impressive,
surprising, and a major breakthrough. Their capabilities are very useful in
certain tasks, such as writing first drafts of documents, code writing
assistance, text summarization, etc. However, their success is limited in
highly scientific domains as they cannot yet reason, plan, or explain due to
their lack of in-depth domain knowledge. This is a problem in domains such as
chemical engineering as they are governed by fundamental laws of physics and
chemistry (and biology), constitutive relations, and highly technical knowledge
about materials, processes, and systems. Although purely data-driven machine
learning has its immediate uses, the long-term success of AI in scientific and
engineering domains would depend on developing hybrid AI systems that use first
principles and technical knowledge effectively. We call these hybrid AI systems
Large Knowledge Models (LKMs), as they will not be limited to only NLP-based
techniques or NLP-like applications. In this paper, we discuss the challenges
and opportunities in developing such systems in chemical engineering. | cs.AI | Quo Vadis ChatGPT? From Large Language Models to Large Knowledge Models |
2024-05-29T22:59:11Z | http://arxiv.org/pdf/2406.00061v1 | http://arxiv.org/abs/2406.00061v1 | Megan Flynn, Alexander Wang, Dean Edward Alvarez, Christopher De Sa, Anil Damle | 2024-05-29T22:59:11Z | cs.LG, cs.AI, cs.CL | We present STAT: a simple algorithm to prune transformer models without any
fine-tuning. STAT eliminates both attention heads and neurons from the network,
while preserving accuracy by calculating a correction to the weights of the
next layer. Each layer block in the network is compressed using a series of
principled matrix factorizations that preserve the network structure. Our
entire algorithm takes minutes to compress BERT, and less than three hours to
compress models with 7B parameters using a single GPU. Using only several
hundred data examples, STAT preserves the output of the network and improves
upon existing gradient-free pruning methods. It is even competitive with
methods that include significant fine-tuning. We demonstrate our method on both
encoder and decoder architectures, including BERT, DistilBERT, and Llama-2
using benchmarks such as GLUE, Squad, WikiText2. | cs.LG | STAT: Shrinking Transformers After Training |
2024-05-29T22:28:46Z | http://arxiv.org/pdf/2406.00060v1 | http://arxiv.org/abs/2406.00060v1 | Congchao Wang, Sean Augenstein, Keith Rush, Wittawat Jitkrittum, Harikrishna Narasimhan, Ankit Singh Rawat, Aditya Krishna Menon, Alec Go | 2024-05-29T22:28:46Z | cs.CL, cs.LG | Reducing serving cost and latency is a fundamental concern for the deployment
of language models (LMs) in business applications. To address this, cascades of
LMs offer an effective solution that conditionally employ smaller models for
simpler queries. Cascaded systems are typically built with independently
trained models, neglecting the advantages of considering inference-time
interactions of the cascaded LMs during training. In this paper, we present
cascade-aware training(CAT), an approach to optimizing the overall quality-cost
performance tradeoff of a cascade of LMs. We achieve inference-time benefits by
training the small LM with awareness of its place in a cascade and downstream
capabilities. We demonstrate the value of the proposed method with over 60 LM
tasks of the SuperGLUE, WMT22, and FLAN2021 datasets. | cs.CL | Cascade-Aware Training of Language Models |
2024-05-29T22:26:26Z | http://arxiv.org/pdf/2405.19550v1 | http://arxiv.org/abs/2405.19550v1 | Ryan Greenblatt, Fabien Roger, Dmitrii Krasheninnikov, David Krueger | 2024-05-29T22:26:26Z | cs.LG, cs.CL | To determine the safety of large language models (LLMs), AI developers must
be able to assess their dangerous capabilities. But simple prompting strategies
often fail to elicit an LLM's full capabilities. One way to elicit capabilities
more robustly is to fine-tune the LLM to complete the task. In this paper, we
investigate the conditions under which fine-tuning-based elicitation suffices
to elicit capabilities. To do this, we introduce password-locked models, LLMs
fine-tuned such that some of their capabilities are deliberately hidden.
Specifically, these LLMs are trained to exhibit these capabilities only when a
password is present in the prompt, and to imitate a much weaker LLM otherwise.
Password-locked models enable a novel method of evaluating capabilities
elicitation methods, by testing whether these password-locked capabilities can
be elicited without using the password. We find that a few high-quality
demonstrations are often sufficient to fully elicit password-locked
capabilities. More surprisingly, fine-tuning can elicit other capabilities that
have been locked using the same password, or even different passwords.
Furthermore, when only evaluations, and not demonstrations, are available,
approaches like reinforcement learning are still often able to elicit
capabilities. Overall, our findings suggest that fine-tuning is an effective
method of eliciting hidden capabilities of current models, but may be
unreliable when high-quality demonstrations are not available, e.g. as may be
the case when models' (hidden) capabilities exceed those of human
demonstrators. | cs.LG | Stress-Testing Capability Elicitation With Password-Locked Models |
2024-05-29T22:12:52Z | http://arxiv.org/pdf/2405.19544v1 | http://arxiv.org/abs/2405.19544v1 | Xinmeng Huang, Shuo Li, Edgar Dobriban, Osbert Bastani, Hamed Hassani, Dongsheng Ding | 2024-05-29T22:12:52Z | cs.AI, cs.CL, cs.LG, math.OC, stat.ML | The growing safety concerns surrounding Large Language Models (LLMs) raise an
urgent need to align them with diverse human preferences to simultaneously
enhance their helpfulness and safety. A promising approach is to enforce safety
constraints through Reinforcement Learning from Human Feedback (RLHF). For such
constrained RLHF, common Lagrangian-based primal-dual policy optimization
methods are computationally expensive and often unstable. This paper presents a
dualization perspective that reduces constrained alignment to an equivalent
unconstrained alignment problem. We do so by pre-optimizing a smooth and convex
dual function that has a closed form. This shortcut eliminates the need for
cumbersome primal-dual policy iterations, thus greatly reducing the
computational burden and improving training stability. Our strategy leads to
two practical algorithms in model-based and preference-based scenarios (MoCAN
and PeCAN, respectively). A broad range of experiments demonstrate the
effectiveness of our methods. | cs.AI | One-Shot Safety Alignment for Large Language Models via Optimal Dualization |
2024-05-29T21:48:56Z | http://arxiv.org/pdf/2405.19538v2 | http://arxiv.org/abs/2405.19538v2 | Pierre Chambon, Jean-Benoit Delbrouck, Thomas Sounack, Shih-Cheng Huang, Zhihong Chen, Maya Varma, Steven QH Truong, Chu The Chuong, Curtis P. Langlotz | 2024-06-03T19:14:12Z | cs.CL, cs.AI, cs.CV, cs.LG | Since the release of the original CheXpert paper five years ago, CheXpert has
become one of the most widely used and cited clinical AI datasets. The
emergence of vision language models has sparked an increase in demands for
sharing reports linked to CheXpert images, along with a growing interest among
AI fairness researchers in obtaining demographic data. To address this,
CheXpert Plus serves as a new collection of radiology data sources, made
publicly available to enhance the scaling, performance, robustness, and
fairness of models for all subsequent machine learning tasks in the field of
radiology. CheXpert Plus is the largest text dataset publicly released in
radiology, with a total of 36 million text tokens, including 13 million
impression tokens. To the best of our knowledge, it represents the largest text
de-identification effort in radiology, with almost 1 million PHI spans
anonymized. It is only the second time that a large-scale English paired
dataset has been released in radiology, thereby enabling, for the first time,
cross-institution training at scale. All reports are paired with high-quality
images in DICOM format, along with numerous image and patient metadata covering
various clinical and socio-economic groups, as well as many pathology labels
and RadGraph annotations. We hope this dataset will boost research for AI
models that can further assist radiologists and help improve medical care. Data
is available at the following URL:
https://stanfordaimi.azurewebsites.net/datasets/5158c524-d3ab-4e02-96e9-6ee9efc110a1
Models are available at the following URL:
https://github.com/Stanford-AIMI/chexpert-plus | cs.CL | CheXpert Plus: Augmenting a Large Chest X-ray Dataset with Text Radiology Reports, Patient Demographics and Additional Image Formats |
2024-05-29T21:29:44Z | http://arxiv.org/pdf/2405.19534v1 | http://arxiv.org/abs/2405.19534v1 | Angelica Chen, Sadhika Malladi, Lily H. Zhang, Xinyi Chen, Qiuyi Zhang, Rajesh Ranganath, Kyunghyun Cho | 2024-05-29T21:29:44Z | cs.LG, cs.AI, cs.CL | Preference learning algorithms (e.g., RLHF and DPO) are frequently used to
steer LLMs to produce generations that are more preferred by humans, but our
understanding of their inner workings is still limited. In this work, we study
the conventional wisdom that preference learning trains models to assign higher
likelihoods to more preferred outputs than less preferred outputs, measured via
$\textit{ranking accuracy}$. Surprisingly, we find that most state-of-the-art
preference-tuned models achieve a ranking accuracy of less than 60% on common
preference datasets. We furthermore derive the $\textit{idealized ranking
accuracy}$ that a preference-tuned LLM would achieve if it optimized the DPO or
RLHF objective perfectly. We demonstrate that existing models exhibit a
significant $\textit{alignment gap}$ -- $\textit{i.e.}$, a gap between the
observed and idealized ranking accuracies. We attribute this discrepancy to the
DPO objective, which is empirically and theoretically ill-suited to fix even
mild ranking errors in the reference model, and derive a simple and efficient
formula for quantifying the difficulty of learning a given preference
datapoint. Finally, we demonstrate that ranking accuracy strongly correlates
with the empirically popular win rate metric when the model is close to the
reference model used in the objective, shedding further light on the
differences between on-policy (e.g., RLHF) and off-policy (e.g., DPO)
preference learning algorithms. | cs.LG | Preference Learning Algorithms Do Not Learn Preference Rankings |
2024-05-29T21:24:15Z | http://arxiv.org/pdf/2406.00059v2 | http://arxiv.org/abs/2406.00059v2 | Yechen Xu, Xinhao Kong, Tingjun Chen, Danyang Zhuo | 2024-06-04T19:00:36Z | cs.CL, cs.DC, cs.LG | The complexity of large language model (LLM) serving workloads has
substantially increased due to the integration with external tool invocations,
such as ChatGPT plugins. In this paper, we identify a new opportunity for
efficient LLM serving for requests that trigger tools: tool partial execution
alongside LLM decoding. To this end, we design Conveyor, an efficient LLM
serving system optimized for handling requests involving external tools. We
introduce a novel interface for tool developers to expose partial execution
opportunities to the LLM serving system and a request scheduler that
facilitates partial tool execution. Our results demonstrate that tool partial
execution can improve request completion latency by up to 38.8%. | cs.CL | Conveyor: Efficient Tool-aware LLM Serving with Tool Partial Execution |
2024-05-29T20:56:52Z | http://arxiv.org/pdf/2405.19519v1 | http://arxiv.org/abs/2405.19519v1 | Sudeshna Das, Yao Ge, Yuting Guo, Swati Rajwal, JaMor Hairston, Jeanne Powell, Drew Walker, Snigdha Peddireddy, Sahithi Lakamana, Selen Bozkurt, Matthew Reyna, Reza Sameni, Yunyu Xiao, Sangmi Kim, Rasheeta Chandler, Natalie Hernandez, Danielle Mowery, Rachel Wightman, Jennifer Love, Anthony Spadaro, Jeanmarie Perrone, Abeed Sarker | 2024-05-29T20:56:52Z | cs.CL, cs.AI | Retrieval augmented generation (RAG) provides the capability to constrain
generative model outputs, and mitigate the possibility of hallucination, by
providing relevant in-context text. The number of tokens a generative large
language model (LLM) can incorporate as context is finite, thus limiting the
volume of knowledge from which to generate an answer. We propose a two-layer
RAG framework for query-focused answer generation and evaluate a
proof-of-concept for this framework in the context of query-focused summary
generation from social media forums, focusing on emerging drug-related
information. The evaluations demonstrate the effectiveness of the two-layer
framework in resource constrained settings to enable researchers in obtaining
near real-time data from users. | cs.CL | Two-layer retrieval augmented generation framework for low-resource medical question-answering: proof of concept using Reddit data |
2024-05-29T20:05:46Z | http://arxiv.org/pdf/2405.19487v1 | http://arxiv.org/abs/2405.19487v1 | Peng Wang, Songshuo Lu, Yaohua Tang, Sijie Yan, Yuanjun Xiong, Wei Xia | 2024-05-29T20:05:46Z | cs.CL | We present a generative dialogue system capable of operating in a full-duplex
manner, allowing for seamless interaction. It is based on a large language
model (LLM) carefully aligned to be aware of a perception module, a motor
function module, and the concept of a simple finite state machine (called
neural FSM) with two states. The perception and motor function modules operate
simultaneously, allowing the system to simultaneously speak and listen to the
user. The LLM generates textual tokens for inquiry responses and makes
autonomous decisions to start responding to, wait for, or interrupt the user by
emitting control tokens to the neural FSM. All these tasks of the LLM are
carried out as next token prediction on a serialized view of the dialogue in
real-time. In automatic quality evaluations simulating real-life interaction,
the proposed system reduces the average conversation response latency by more
than 3 folds compared with LLM-based half-duplex dialogue systems while
responding within less than 500 milliseconds in more than 50% of evaluated
interactions. Running a LLM with only 8 billion parameters, our system exhibits
a 8% higher interruption precision rate than the best available commercial LLM
for voice-based dialogue. | cs.CL | A Full-duplex Speech Dialogue Scheme Based On Large Language Models |
2024-05-29T19:21:49Z | http://arxiv.org/pdf/2405.19462v2 | http://arxiv.org/abs/2405.19462v2 | Everlyn Asiko Chimoto, Jay Gala, Orevaoghene Ahia, Julia Kreutzer, Bruce A. Bassett, Sara Hooker | 2024-06-21T12:30:36Z | cs.CL | Neural Machine Translation models are extremely data and compute-hungry.
However, not all data points contribute equally to model training and
generalization. Data pruning to remove the low-value data points has the
benefit of drastically reducing the compute budget without significant drop in
model performance. In this paper, we propose a new data pruning technique:
Checkpoints Across Time (CAT), that leverages early model training dynamics to
identify the most relevant data points for model performance. We benchmark CAT
against several data pruning techniques including COMET-QE, LASER and LaBSE. We
find that CAT outperforms the benchmarks on Indo-European languages on multiple
test sets. When applied to English-German, English-French and English-Swahili
translation tasks, CAT achieves comparable performance to using the full
dataset, while pruning up to 50% of training data. We inspect the data points
that CAT selects and find that it tends to favour longer sentences and
sentences with unique or rare words. | cs.CL | Critical Learning Periods: Leveraging Early Training Dynamics for Efficient Data Pruning |
2024-05-29T18:19:46Z | http://arxiv.org/pdf/2406.00057v2 | http://arxiv.org/abs/2406.00057v2 | Nick Alonso, Tomás Figliolia, Anthony Ndirango, Beren Millidge | 2024-06-04T18:01:03Z | cs.CL, cs.LG | There has recently been growing interest in conversational agents with
long-term memory which has led to the rapid development of language models that
use retrieval-augmented generation (RAG). Until recently, most work on RAG has
focused on information retrieval from large databases of texts, like Wikipedia,
rather than information from long-form conversations. In this paper, we argue
that effective retrieval from long-form conversational data faces two unique
problems compared to static database retrieval: 1) time/event-based queries,
which requires the model to retrieve information about previous conversations
based on time or the order of a conversational event (e.g., the third
conversation on Tuesday), and 2) ambiguous queries that require surrounding
conversational context to understand. To better develop RAG-based agents that
can deal with these challenges, we generate a new dataset of ambiguous and
time-based questions that build upon a recent dataset of long-form, simulated
conversations, and demonstrate that standard RAG based approaches handle such
questions poorly. We then develop a novel retrieval model which combines
chained-of-table search methods, standard vector-database retrieval, and a
prompting method to disambiguate queries, and demonstrate that this approach
substantially improves over current methods at solving these tasks. We believe
that this new dataset and more advanced RAG agent can act as a key benchmark
and stepping stone towards effective memory augmented conversational agents
that can be used in a wide variety of AI applications. | cs.CL | Toward Conversational Agents with Context and Time Sensitive Long-term Memory |
2024-05-29T18:16:32Z | http://arxiv.org/pdf/2405.19433v1 | http://arxiv.org/abs/2405.19433v1 | Yupei Wang, Renfen Hu, Zhe Zhao | 2024-05-29T18:16:32Z | cs.CL | While current automated essay scoring (AES) methods show high agreement with
human raters, their scoring mechanisms are not fully explored. Our proposed
method, using counterfactual intervention assisted by Large Language Models
(LLMs), reveals that when scoring essays, BERT-like models primarily focus on
sentence-level features, while LLMs are attuned to conventions, language
complexity, as well as organization, indicating a more comprehensive alignment
with scoring rubrics. Moreover, LLMs can discern counterfactual interventions
during feedback. Our approach improves understanding of neural AES methods and
can also apply to other domains seeking transparency in model-driven decisions.
The codes and data will be released at GitHub. | cs.CL | Beyond Agreement: Diagnosing the Rationale Alignment of Automated Essay Scoring Methods based on Linguistically-informed Counterfactuals |
2024-05-29T18:09:35Z | http://arxiv.org/pdf/2405.19426v2 | http://arxiv.org/abs/2405.19426v2 | Mithilesh Vaidya, Binaya Kumar Sahoo, Preeti Rao | 2024-06-01T14:16:42Z | cs.CL, cs.SD, eess.AS | Reading fluency assessment is a critical component of literacy programmes,
serving to guide and monitor early education interventions. Given the resource
intensive nature of the exercise when conducted by teachers, the development of
automatic tools that can operate on audio recordings of oral reading is
attractive as an objective and highly scalable solution. Multiple complex
aspects such as accuracy, rate and expressiveness underlie human judgements of
reading fluency. In this work, we investigate end-to-end modeling on a training
dataset of children's audio recordings of story texts labeled by human experts.
The pre-trained wav2vec2.0 model is adopted due its potential to alleviate the
challenges from the limited amount of labeled data. We report the performance
of a number of system variations on the relevant measures, and also probe the
learned embeddings for lexical and acoustic-prosodic features known to be
important to the perception of reading fluency. | cs.CL | Deep Learning for Assessment of Oral Reading Fluency |
2024-05-29T18:08:37Z | http://arxiv.org/pdf/2405.19425v1 | http://arxiv.org/abs/2405.19425v1 | Linxin Song, Jiale Liu, Jieyu Zhang, Shaokun Zhang, Ao Luo, Shijian Wang, Qingyun Wu, Chi Wang | 2024-05-29T18:08:37Z | cs.CL | Leveraging multiple large language model (LLM) agents has shown to be a
promising approach for tackling complex tasks, while the effective design of
multiple agents for a particular application remains an art. It is thus
intriguing to answer a critical question: Given a task, how can we build a team
of LLM agents to solve it effectively? Our new adaptive team-building paradigm
offers a flexible solution, realized through a novel agent design named Captain
Agent. It dynamically forms and manages teams for each step of a task-solving
process, utilizing nested group conversations and reflection to ensure diverse
expertise and prevent stereotypical outputs. It allows for a flexible yet
structured approach to problem-solving and can help reduce redundancy and
enhance output diversity. A comprehensive evaluation across six real-world
scenarios demonstrates that Captain Agent significantly outperforms existing
multi-agent methods with 21.94% improvement in average accuracy, providing
outstanding performance without requiring task-specific prompt engineering. | cs.CL | Adaptive In-conversation Team Building for Language Model Agents |
2024-05-29T17:59:58Z | http://arxiv.org/pdf/2405.19335v1 | http://arxiv.org/abs/2405.19335v1 | Hanrong Ye, De-An Huang, Yao Lu, Zhiding Yu, Wei Ping, Andrew Tao, Jan Kautz, Song Han, Dan Xu, Pavlo Molchanov, Hongxu Yin | 2024-05-29T17:59:58Z | cs.CV, cs.CL, cs.LG | We introduce X-VILA, an omni-modality model designed to extend the
capabilities of large language models (LLMs) by incorporating image, video, and
audio modalities. By aligning modality-specific encoders with LLM inputs and
diffusion decoders with LLM outputs, X-VILA achieves cross-modality
understanding, reasoning, and generation. To facilitate this cross-modality
alignment, we curate an effective interleaved any-to-any modality
instruction-following dataset. Furthermore, we identify a significant problem
with the current cross-modality alignment method, which results in visual
information loss. To address the issue, we propose a visual alignment mechanism
with a visual embedding highway module. We then introduce a resource-efficient
recipe for training X-VILA, that exhibits proficiency in any-to-any modality
conversation, surpassing previous approaches by large margins. X-VILA also
showcases emergent properties across modalities even in the absence of similar
training data. The project will be made open-source. | cs.CV | X-VILA: Cross-Modality Alignment for Large Language Model |
2024-05-29T17:59:20Z | http://arxiv.org/pdf/2405.19334v2 | http://arxiv.org/abs/2405.19334v2 | Yingqing He, Zhaoyang Liu, Jingye Chen, Zeyue Tian, Hongyu Liu, Xiaowei Chi, Runtao Liu, Ruibin Yuan, Yazhou Xing, Wenhai Wang, Jifeng Dai, Yong Zhang, Wei Xue, Qifeng Liu, Yike Guo, Qifeng Chen | 2024-06-09T11:34:12Z | cs.AI, cs.CL, cs.CV, cs.MM, cs.SD | With the recent advancement in large language models (LLMs), there is a
growing interest in combining LLMs with multimodal learning. Previous surveys
of multimodal large language models (MLLMs) mainly focus on multimodal
understanding. This survey elaborates on multimodal generation and editing
across various domains, comprising image, video, 3D, and audio. Specifically,
we summarize the notable advancements with milestone works in these fields and
categorize these studies into LLM-based and CLIP/T5-based methods. Then, we
summarize the various roles of LLMs in multimodal generation and exhaustively
investigate the critical technical components behind these methods and the
multimodal datasets utilized in these studies. Additionally, we dig into
tool-augmented multimodal agents that can leverage existing generative models
for human-computer interaction. Lastly, we discuss the advancements in the
generative AI safety field, investigate emerging applications, and discuss
future prospects. Our work provides a systematic and insightful overview of
multimodal generation and processing, which is expected to advance the
development of Artificial Intelligence for Generative Content (AIGC) and world
models. A curated list of all related papers can be found at
https://github.com/YingqingHe/Awesome-LLMs-meet-Multimodal-Generation | cs.AI | LLMs Meet Multimodal Generation and Editing: A Survey |
2024-05-29T17:57:16Z | http://arxiv.org/pdf/2405.19327v3 | http://arxiv.org/abs/2405.19327v3 | Ge Zhang, Scott Qu, Jiaheng Liu, Chenchen Zhang, Chenghua Lin, Chou Leuang Yu, Danny Pan, Esther Cheng, Jie Liu, Qunshu Lin, Raven Yuan, Tuney Zheng, Wei Pang, Xinrun Du, Yiming Liang, Yinghao Ma, Yizhi Li, Ziyang Ma, Bill Lin, Emmanouil Benetos, Huan Yang, Junting Zhou, Kaijing Ma, Minghao Liu, Morry Niu, Noah Wang, Quehry Que, Ruibo Liu, Sine Liu, Shawn Guo, Soren Gao, Wangchunshu Zhou, Xinyue Zhang, Yizhi Zhou, Yubo Wang, Yuelin Bai, Yuhan Zhang, Yuxiang Zhang, Zenith Wang, Zhenzhu Yang, Zijian Zhao, Jiajun Zhang, Wanli Ouyang, Wenhao Huang, Wenhu Chen | 2024-06-02T06:09:49Z | cs.CL, cs.AI, cs.LG | Large Language Models (LLMs) have made great strides in recent years to
achieve unprecedented performance across different tasks. However, due to
commercial interest, the most competitive models like GPT, Gemini, and Claude
have been gated behind proprietary interfaces without disclosing the training
details. Recently, many institutions have open-sourced several strong LLMs like
LLaMA-3, comparable to existing closed-source LLMs. However, only the model's
weights are provided with most details (e.g., intermediate checkpoints,
pre-training corpus, and training code, etc.) being undisclosed. To improve the
transparency of LLMs, the research community has formed to open-source truly
open LLMs (e.g., Pythia, Amber, OLMo), where more details (e.g., pre-training
corpus and training code) are being provided. These models have greatly
advanced the scientific study of these large models including their strengths,
weaknesses, biases and risks. However, we observe that the existing truly open
LLMs on reasoning, knowledge, and coding tasks are still inferior to existing
state-of-the-art LLMs with similar model sizes. To this end, we open-source
MAP-Neo, a highly capable and transparent bilingual language model with 7B
parameters trained from scratch on 4.5T high-quality tokens. Our MAP-Neo is the
first fully open-sourced bilingual LLM with comparable performance compared to
existing state-of-the-art LLMs. Moreover, we open-source all details to
reproduce our MAP-Neo, where the cleaned pre-training corpus, data cleaning
pipeline, checkpoints, and well-optimized training/evaluation framework are
provided. Finally, we hope our MAP-Neo will enhance and strengthen the open
research community and inspire more innovations and creativities to facilitate
the further improvements of LLMs. | cs.CL | MAP-Neo: Highly Capable and Transparent Bilingual Large Language Model Series |
2024-05-29T17:55:03Z | http://arxiv.org/pdf/2405.19325v2 | http://arxiv.org/abs/2405.19325v2 | Minghan Li, Xilun Chen, Ari Holtzman, Beidi Chen, Jimmy Lin, Wen-tau Yih, Xi Victoria Lin | 2024-05-31T01:41:49Z | cs.CL | Large language models (LLMs) often hallucinate and lack the ability to
provide attribution for their generations. Semi-parametric LMs, such as kNN-LM,
approach these limitations by refining the output of an LM for a given prompt
using its nearest neighbor matches in a non-parametric data store. However,
these models often exhibit slow inference speeds and produce non-fluent texts.
In this paper, we introduce Nearest Neighbor Speculative Decoding (NEST), a
novel semi-parametric language modeling approach that is capable of
incorporating real-world text spans of arbitrary length into the LM generations
and providing attribution to their sources. NEST performs token-level retrieval
at each inference step to compute a semi-parametric mixture distribution and
identify promising span continuations in a corpus. It then uses an approximate
speculative decoding procedure that accepts a prefix of the retrieved span or
generates a new token. NEST significantly enhances the generation quality and
attribution rate of the base LM across a variety of knowledge-intensive tasks,
surpassing the conventional kNN-LM method and performing competitively with
in-context retrieval augmentation. In addition, NEST substantially improves the
generation speed, achieving a 1.8x speedup in inference time when applied to
Llama-2-Chat 70B. | cs.CL | Nearest Neighbor Speculative Decoding for LLM Generation and Attribution |
2024-05-29T17:54:22Z | http://arxiv.org/pdf/2405.19323v1 | http://arxiv.org/abs/2405.19323v1 | Mingmeng Geng, Sihong He, Roberto Trotta | 2024-05-29T17:54:22Z | cs.CL, cs.AI, cs.CY, cs.LG | Do large language models (LLMs) have their own worldviews and personality
tendencies? Simulations in which an LLM was asked to answer subjective
questions were conducted more than 1 million times. Comparison of the responses
from different LLMs with real data from the European Social Survey (ESS)
suggests that the effect of prompts on bias and variability is fundamental,
highlighting major cultural, age, and gender biases. Methods for measuring the
difference between LLMs and survey data are discussed, such as calculating
weighted means and a new proposed measure inspired by Jaccard similarity. We
conclude that it is important to analyze the robustness and variability of
prompts before using LLMs to model individual decisions or collective behavior,
as their imitation abilities are approximate at best. | cs.CL | Are Large Language Models Chameleons? |
2024-05-29T17:39:48Z | http://arxiv.org/pdf/2405.19316v1 | http://arxiv.org/abs/2405.19316v1 | Adam Fisch, Jacob Eisenstein, Vicky Zayats, Alekh Agarwal, Ahmad Beirami, Chirag Nagpal, Pete Shaw, Jonathan Berant | 2024-05-29T17:39:48Z | cs.LG, cs.CL | Language model (LM) post-training (or alignment) involves maximizing a reward
function that is derived from preference annotations. Direct Preference
Optimization (DPO) is a popular offline alignment method that trains a policy
directly on preference data without the need to train a reward model or apply
reinforcement learning. However, typical preference datasets have only a
single, or at most a few, annotation per preference pair, which causes DPO to
overconfidently assign rewards that trend towards infinite magnitude. This
frequently leads to degenerate policies, sometimes causing even the
probabilities of the preferred generations to go to zero. In this work, we
analyze this phenomenon and propose distillation to get a better proxy for the
true preference distribution over generation pairs: we train the LM to produce
probabilities that match the distribution induced by a reward model trained on
the preference data. Moreover, to account for uncertainty in the reward model
we are distilling from, we optimize against a family of reward models that, as
a whole, is likely to include at least one reasonable proxy for the preference
distribution. Our results show that distilling from such a family of reward
models leads to improved robustness to distribution shift in preference
annotations, while preserving the simple supervised nature of DPO. | cs.LG | Robust Preference Optimization through Reward Model Distillation |
2024-05-29T17:39:42Z | http://arxiv.org/pdf/2405.19315v2 | http://arxiv.org/abs/2405.19315v2 | Wenbo Hu, Zi-Yi Dou, Liunian Harold Li, Amita Kamath, Nanyun Peng, Kai-Wei Chang | 2024-06-07T03:39:04Z | cs.CV, cs.CL, cs.LG | Large Vision-Language Models (LVLMs) typically encode an image into a fixed
number of visual tokens (e.g., 576) and process these tokens with a language
model. Despite their strong performance, LVLMs face challenges in adapting to
varying computational constraints. This raises the question: can we achieve
flexibility in the number of visual tokens to suit different tasks and
computational resources? We answer this with an emphatic yes. Inspired by
Matryoshka Representation Learning, we introduce the Matryoshka Query
Transformer (MQT), capable of encoding an image into m visual tokens during
inference, where m can be any number up to a predefined maximum. This is
achieved by employing a query transformer with M latent query tokens to
compress the visual embeddings. During each training step, we randomly select m
<= M latent query tokens and train the model using only these first m tokens,
discarding the rest. Combining MQT with LLaVA, we train a single model once,
and flexibly and drastically reduce the number of inference-time visual tokens
while maintaining similar or better performance compared to training
independent models for each number of tokens. Our model, MQT-LLAVA, matches
LLaVA-1.5 performance across 11 benchmarks using a maximum of 256 tokens
instead of LLaVA's fixed 576. Reducing to 16 tokens (8x less TFLOPs) only
sacrifices the performance by 2.4 points on MMBench. On certain tasks such as
ScienceQA and MMMU, we can even go down to only 2 visual tokens with
performance drops of just 3% and 6% each. Our exploration of the trade-off
between the accuracy and computational cost brought about by the number of
visual tokens facilitates future research to achieve the best of both worlds. | cs.CV | Matryoshka Query Transformer for Large Vision-Language Models |
2024-05-29T17:37:14Z | http://arxiv.org/pdf/2405.19313v1 | http://arxiv.org/abs/2405.19313v1 | Jian-Qiao Zhu, Haijiang Yan, Thomas L. Griffiths | 2024-05-29T17:37:14Z | cs.AI, cs.CL, econ.GN, q-fin.EC | The observed similarities in the behavior of humans and Large Language Models
(LLMs) have prompted researchers to consider the potential of using LLMs as
models of human cognition. However, several significant challenges must be
addressed before LLMs can be legitimately regarded as cognitive models. For
instance, LLMs are trained on far more data than humans typically encounter,
and may have been directly trained on human data in specific cognitive tasks or
aligned with human preferences. Consequently, the origins of these behavioral
similarities are not well understood. In this paper, we propose a novel way to
enhance the utility of LLMs as cognitive models. This approach involves (i)
leveraging computationally equivalent tasks that both an LLM and a rational
agent need to master for solving a cognitive problem and (ii) examining the
specific task distributions required for an LLM to exhibit human-like
behaviors. We apply this approach to decision-making -- specifically risky and
intertemporal choice -- where the key computationally equivalent task is the
arithmetic of expected value calculations. We show that an LLM pretrained on an
ecologically valid arithmetic dataset, which we call Arithmetic-GPT, predicts
human behavior better than many traditional cognitive models. Pretraining LLMs
on ecologically valid arithmetic datasets is sufficient to produce a strong
correspondence between these models and human decision-making. Our results also
suggest that LLMs used as cognitive models should be carefully investigated via
ablation studies of the pretraining data. | cs.AI | Language Models Trained to do Arithmetic Predict Human Risky and Intertemporal Choice |
2024-05-29T17:26:52Z | http://arxiv.org/pdf/2405.19299v1 | http://arxiv.org/abs/2405.19299v1 | Xueyao Sun, Kaize Shi, Haoran Tang, Guandong Xu, Qing Li | 2024-05-29T17:26:52Z | cs.CL | Large language models (LLMs) can elicit social bias during generations,
especially when inference with toxic prompts. Controlling the sensitive
attributes in generation encounters challenges in data distribution,
generalizability, and efficiency. Specifically, fine-tuning and retrieval
demand extensive unbiased corpus, while direct prompting requires meticulously
curated instructions for correcting the output in multiple rounds of thoughts
but poses challenges on memory and inference latency. In this work, we propose
the Expert-Guided Extinction of Toxic Tokens for Debiased Generation (EXPOSED)
to eliminate the undesired harmful outputs for LLMs without the aforementioned
requirements. EXPOSED constructs a debiasing expert based on the abundant toxic
corpus to expose and elicit the potentially dangerous tokens. It then processes
the output to the LLMs and constructs a fair distribution by suppressing and
attenuating the toxic tokens. EXPOSED is evaluated on fairness benchmarks over
three LLM families. Extensive experiments demonstrate that compared with other
baselines, the proposed EXPOSED significantly reduces the potential social bias
while balancing fairness and generation performance. | cs.CL | Expert-Guided Extinction of Toxic Tokens for Debiased Generation |
2024-05-29T17:19:04Z | http://arxiv.org/pdf/2405.19290v2 | http://arxiv.org/abs/2405.19290v2 | Langlin Huang, Yang Feng | 2024-06-07T08:34:57Z | cs.CL | Subword tokenization is a common method for vocabulary building in Neural
Machine Translation (NMT) models. However, increasingly complex tasks have
revealed its disadvantages. First, a vocabulary cannot be modified once it is
learned, making it hard to adapt to new words. Second, in multilingual
translation, the imbalance in data volumes across different languages spreads
to the vocabulary, exacerbating translations involving low-resource languages.
While byte-based tokenization addresses these issues, byte-based models
struggle with the low information density inherent in UTF-8 byte sequences.
Previous works enhance token semantics through local contextualization but fail
to select an appropriate contextualizing scope based on the input.
Consequently, we propose the Multi-Scale Contextualization (MSC) method, which
learns contextualized information of varying scales across different hidden
state dimensions. It then leverages the attention module to dynamically
integrate the multi-scale contextualized information. Experiments show that MSC
significantly outperforms subword-based and other byte-based methods in both
multilingual and out-of-domain scenarios. Code can be found in
https://github.com/ictnlp/Multiscale-Contextualization. | cs.CL | Integrating Multi-scale Contextualized Information for Byte-based Neural Machine Translation |
2024-05-29T17:17:22Z | http://arxiv.org/pdf/2405.19285v1 | http://arxiv.org/abs/2405.19285v1 | Michael Regan, Shira Wein, George Baker, Emilio Monti | 2024-05-29T17:17:22Z | cs.CL | Abstract Meaning Representation (AMR) is a semantic formalism that captures
the core meaning of an utterance. There has been substantial work developing
AMR corpora in English and more recently across languages, though the limited
size of existing datasets and the cost of collecting more annotations are
prohibitive. With both engineering and scientific questions in mind, we
introduce MASSIVE-AMR, a dataset with more than 84,000 text-to-graph
annotations, currently the largest and most diverse of its kind: AMR graphs for
1,685 information-seeking utterances mapped to 50+ typologically diverse
languages. We describe how we built our resource and its unique features before
reporting on experiments using large language models for multilingual AMR and
SPARQL parsing as well as applying AMRs for hallucination detection in the
context of knowledge base question answering, with results shedding light on
persistent issues using LLMs for structured parsing. | cs.CL | MASSIVE Multilingual Abstract Meaning Representation: A Dataset and Baselines for Hallucination Detection |
2024-05-29T16:59:38Z | http://arxiv.org/pdf/2405.19266v2 | http://arxiv.org/abs/2405.19266v2 | Dingkang Yang, Jinjie Wei, Dongling Xiao, Shunli Wang, Tong Wu, Gang Li, Mingcheng Li, Shuaibing Wang, Jiawei Chen, Yue Jiang, Qingyao Xu, Ke Li, Peng Zhai, Lihua Zhang | 2024-06-03T15:27:10Z | cs.CL | Developing intelligent pediatric consultation systems offers promising
prospects for improving diagnostic efficiency, especially in China, where
healthcare resources are scarce. Despite recent advances in Large Language
Models (LLMs) for Chinese medicine, their performance is sub-optimal in
pediatric applications due to inadequate instruction data and vulnerable
training procedures. To address the above issues, this paper builds PedCorpus,
a high-quality dataset of over 300,000 multi-task instructions from pediatric
textbooks, guidelines, and knowledge graph resources to fulfil diverse
diagnostic demands. Upon well-designed PedCorpus, we propose PediatricsGPT, the
first Chinese pediatric LLM assistant built on a systematic and robust training
pipeline. In the continuous pre-training phase, we introduce a hybrid
instruction pre-training mechanism to mitigate the internal-injected knowledge
inconsistency of LLMs for medical domain adaptation. Immediately, the
full-parameter Supervised Fine-Tuning (SFT) is utilized to incorporate the
general medical knowledge schema into the models. After that, we devise a
direct following preference optimization to enhance the generation of
pediatrician-like humanistic responses. In the parameter-efficient secondary
SFT phase, a mixture of universal-specific experts strategy is presented to
resolve the competency conflict between medical generalist and pediatric
expertise mastery. Extensive results based on the metrics, GPT-4, and doctor
evaluations on distinct doctor downstream tasks show that PediatricsGPT
consistently outperforms previous Chinese medical LLMs. Our model and dataset
will be open-source for community development. | cs.CL | PediatricsGPT: Large Language Models as Chinese Medical Assistants for Pediatric Applications |
2024-05-29T16:57:33Z | http://arxiv.org/pdf/2405.19265v1 | http://arxiv.org/abs/2405.19265v1 | Zifan Song, Yudong Wang, Wenwei Zhang, Kuikun Liu, Chengqi Lyu, Demin Song, Qipeng Guo, Hang Yan, Dahua Lin, Kai Chen, Cairong Zhao | 2024-05-29T16:57:33Z | cs.CL | Open-source Large Language Models (LLMs) and their specialized variants,
particularly Code LLMs, have recently delivered impressive performance.
However, previous Code LLMs are typically fine-tuned on single-source data with
limited quality and diversity, which may insufficiently elicit the potential of
pre-trained Code LLMs. In this paper, we present AlchemistCoder, a series of
Code LLMs with enhanced code generation and generalization capabilities
fine-tuned on multi-source data. To achieve this, we pioneer to unveil inherent
conflicts among the various styles and qualities in multi-source code corpora
and introduce data-specific prompts with hindsight relabeling, termed
AlchemistPrompts, to harmonize different data sources and instruction-response
pairs. Additionally, we propose incorporating the data construction process
into the fine-tuning data as code comprehension tasks, including instruction
evolution, data filtering, and code review. Extensive experiments demonstrate
that AlchemistCoder holds a clear lead among all models of the same size
(6.7B/7B) and rivals or even surpasses larger models (15B/33B/70B), showcasing
the efficacy of our method in refining instruction-following capabilities and
advancing the boundaries of code intelligence. | cs.CL | AlchemistCoder: Harmonizing and Eliciting Code Capability by Hindsight Tuning on Multi-source Data |
2024-05-29T16:55:32Z | http://arxiv.org/pdf/2405.19262v1 | http://arxiv.org/abs/2405.19262v1 | Zhanhui Zhou, Zhixuan Liu, Jie Liu, Zhichen Dong, Chao Yang, Yu Qiao | 2024-05-29T16:55:32Z | cs.CL, cs.AI, cs.LG | Large language models are usually fine-tuned to align with human preferences.
However, fine-tuning a large language model can be challenging. In this work,
we introduce $\textit{weak-to-strong search}$, framing the alignment of a large
language model as a test-time greedy search to maximize the log-likelihood
difference between small tuned and untuned models while sampling from the
frozen large model. This method serves both as (i) a compute-efficient model
up-scaling strategy that avoids directly tuning the large model and as (ii) an
instance of weak-to-strong generalization that enhances a strong model with
weak test-time guidance. Empirically, we demonstrate the flexibility of
weak-to-strong search across different tasks. In controlled-sentiment
generation and summarization, we use tuned and untuned $\texttt{gpt2}$s to
effectively improve the alignment of large models without additional training.
Crucially, in a more difficult instruction-following benchmark, AlpacaEval 2.0,
we show that reusing off-the-shelf small model pairs (e.g.,
$\texttt{zephyr-7b-beta}$ and its untuned version) can significantly improve
the length-controlled win rates of both white-box and black-box large models
against $\texttt{gpt-4-turbo}$ (e.g., $34.4 \rightarrow 37.9$ for
$\texttt{Llama-3-70B-Instruct}$ and $16.0 \rightarrow 20.1$ for
$\texttt{gpt-3.5-turbo-instruct}$), despite the small models' low win rates
$\approx 10.0$. | cs.CL | Weak-to-Strong Search: Align Large Language Models via Searching over Small Language Models |
2024-05-29T16:55:08Z | http://arxiv.org/pdf/2405.19261v1 | http://arxiv.org/abs/2405.19261v1 | Harikrishna Narasimhan, Wittawat Jitkrittum, Ankit Singh Rawat, Seungyeon Kim, Neha Gupta, Aditya Krishna Menon, Sanjiv Kumar | 2024-05-29T16:55:08Z | cs.CL, cs.AI, cs.LG | Cascades and speculative decoding are two common approaches to improving
language models' inference efficiency. Both approaches involve interleaving
models of different sizes, but via fundamentally distinct mechanisms: cascades
employ a deferral rule that invokes the larger model only for "hard" inputs,
while speculative decoding uses speculative execution to primarily invoke the
larger model in parallel verification mode. These mechanisms offer different
benefits: empirically, cascades are often capable of yielding better quality
than even the larger model, while theoretically, speculative decoding offers a
guarantee of quality-neutrality. In this paper, we leverage the best of both
these approaches by designing new speculative cascading techniques that
implement their deferral rule through speculative execution. We characterize
the optimal deferral rule for our speculative cascades, and employ a plug-in
approximation to the optimal rule. Through experiments with T5 models on
benchmark language tasks, we show that the proposed approach yields better
cost-quality trade-offs than cascading and speculative decoding baselines. | cs.CL | Faster Cascades via Speculative Decoding |
2024-05-29T16:02:09Z | http://arxiv.org/pdf/2405.19222v2 | http://arxiv.org/abs/2405.19222v2 | Anej Svete, Franz Nowak, Anisha Mohamed Sahabdeen, Ryan Cotterell | 2024-06-18T15:42:32Z | cs.CL | The recent successes and spread of large neural language models (LMs) call
for a thorough understanding of their computational ability. Describing their
computational abilities through LMs' \emph{representational capacity} is a
lively area of research. However, investigation into the representational
capacity of neural LMs has predominantly focused on their ability to
\emph{recognize} formal languages. For example, recurrent neural networks
(RNNs) with Heaviside activations are tightly linked to regular languages,
i.e., languages defined by finite-state automata (FSAs). Such results, however,
fall short of describing the capabilities of RNN \emph{language models} (LMs),
which are definitionally \emph{distributions} over strings. We take a fresh
look at the representational capacity of RNN LMs by connecting them to
\emph{probabilistic} FSAs and demonstrate that RNN LMs with linearly bounded
precision can express arbitrary regular LMs. | cs.CL | Lower Bounds on the Expressivity of Recurrent Neural Language Models |
2024-05-29T16:00:46Z | http://arxiv.org/pdf/2405.19220v3 | http://arxiv.org/abs/2405.19220v3 | Ravil Mussabayev | 2024-06-25T10:41:43Z | cs.CL, cs.AI | The problem of natural language generation, and, more specifically, method
name prediction, faces significant difficulties when proposed models need to be
evaluated on test data. Such a metric would need to consider the versatility
with which a single method can be named, with respect to both semantics and
syntax. Measuring the direct overlap between the predicted and reference (true)
sequences will not be able to capture these subtleties. Other existing
embedding based metrics either do not measure precision and recall or impose
strict unrealistic assumptions on both sequences. To address these issues, we
propose a new metric that, on the one hand, is very simple and lightweight,
and, on the other hand, is able to calculate precision and recall without
resorting to any assumptions while obtaining good performance with respect to
the human judgement. | cs.CL | WRDScore: New Metric for Evaluation of Natural Language Generation Models |
2024-05-29T15:49:09Z | http://arxiv.org/pdf/2405.19209v1 | http://arxiv.org/abs/2405.19209v1 | Ziyang Wang, Shoubin Yu, Elias Stengel-Eskin, Jaehong Yoon, Feng Cheng, Gedas Bertasius, Mohit Bansal | 2024-05-29T15:49:09Z | cs.CV, cs.AI, cs.CL | Video-language understanding tasks have focused on short video clips, often
struggling with long-form video understanding tasks. Recently, many long
video-language understanding approaches have leveraged the reasoning
capabilities of Large Language Models (LLMs) to perform long video QA,
transforming videos into densely sampled frame captions, and asking LLMs to
respond to text queries over captions. However, the frames used for captioning
are often redundant and contain irrelevant information, making dense sampling
inefficient, and ignoring the fact that video QA requires varying levels of
granularity, with some video segments being highly relevant to the question
(needing more fine-grained detail) while others being less relevant. Thus,
these LLM-based approaches are prone to missing information and operate on
large numbers of irrelevant captions, lowering both performance and efficiency.
To address these issues, we introduce VideoTree, a query-adaptive and
hierarchical framework for long-video understanding with LLMs. VideoTree
dynamically extracts query-related information from a video and builds a
tree-based representation for LLM reasoning. First, VideoTree adaptively
selects frames for captioning by iteratively clustering frames based on their
visual features and scoring clusters using their relevance to the query.
Second, it organizes visual clusters into a query-adaptive and hierarchical
tree structure; the tree encodes varying levels of granularity, with higher
resolution on relevant segments. Finally, VideoTree produces an answer by
traversing the tree's keyframes and passing their captions to an LLM answerer.
Our method improves both reasoning accuracy and efficiency compared to existing
methods: VideoTree achieves a 7.0%, 2.2%, and 2.7% accuracy gain over baselines
on the EgoSchema, NExT-QA, and IntentQA benchmarks, respectively, while
reducing inference time by 40%. | cs.CV | VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long Videos |
2024-05-29T15:28:42Z | http://arxiv.org/pdf/2405.19186v1 | http://arxiv.org/abs/2405.19186v1 | Laura Fieback, Jakob Spiegelberg, Hanno Gottschalk | 2024-05-29T15:28:42Z | cs.CV, cs.CL, cs.LG, I.4 | Large Vision Language Models (LVLMs) have shown remarkable capabilities in
multimodal tasks like visual question answering or image captioning. However,
inconsistencies between the visual information and the generated text, a
phenomenon referred to as hallucinations, remain an unsolved problem with
regard to the trustworthiness of LVLMs. To address this problem, recent works
proposed to incorporate computationally costly Large (Vision) Language Models
in order to detect hallucinations on a sentence- or subsentence-level. In this
work, we introduce MetaToken, a lightweight binary classifier to detect
hallucinations on the token-level at negligible cost. Based on a statistical
analysis, we reveal key factors of hallucinations in LVLMs which have been
overseen in previous works. MetaToken can be applied to any open-source LVLM
without any knowledge about ground truth data providing a reliable detection of
hallucinations. We evaluate our method on four state-of-the-art LVLMs
demonstrating the effectiveness of our approach. | cs.CV | MetaToken: Detecting Hallucination in Image Descriptions by Meta Classification |
2024-05-29T14:47:01Z | http://arxiv.org/pdf/2405.19139v1 | http://arxiv.org/abs/2405.19139v1 | Runfeng Lin, Dacheng Xu, Huijiang Wang, Zebiao Chen, Yating Wang, Shouqiang Liu | 2024-05-29T14:47:01Z | cs.CL, cs.AI | When evaluating a learner's knowledge proficiency, the multiple-choice
question is an efficient and widely used format in standardized tests.
Nevertheless, generating these questions, particularly plausible distractors
(incorrect options), poses a considerable challenge. Generally, the distractor
generation can be classified into cloze-style distractor generation (CDG) and
natural questions distractor generation (NQDG). In contrast to the CDG,
utilizing pre-trained language models (PLMs) for NQDG presents three primary
challenges: (1) PLMs are typically trained to generate ``correct'' content,
like answers, while rarely trained to generate ``plausible" content, like
distractors; (2) PLMs often struggle to produce content that aligns well with
specific knowledge and the style of exams; (3) NQDG necessitates the model to
produce longer, context-sensitive, and question-relevant distractors. In this
study, we introduce a fine-tuning framework named DGRC for NQDG in Chinese
multi-choice reading comprehension from authentic examinations. DGRC comprises
three major components: hard chain-of-thought, multi-task learning, and
generation mask patterns. The experiment results demonstrate that DGRC
significantly enhances generation performance, achieving a more than 2.5-fold
improvement in BLEU scores. | cs.CL | DGRC: An Effective Fine-tuning Framework for Distractor Generation in Chinese Multi-choice Reading Comprehension |
2024-05-29T14:14:05Z | http://arxiv.org/pdf/2405.19109v1 | http://arxiv.org/abs/2405.19109v1 | Fangzhi Xu, Qika Lin, Tianzhe Zhao, Jiawei Han, Jun Liu | 2024-05-29T14:14:05Z | cs.CL | Logical reasoning task has attracted great interest since it was proposed.
Faced with such a task, current competitive models, even large language models
(e.g., ChatGPT and PaLM 2), still perform badly. Previous promising LMs
struggle in logical consistency modeling and logical structure perception. To
this end, we model the logical reasoning task by transforming each logical
sample into reasoning paths and propose an architecture \textbf{PathReasoner}.
It addresses the task from the views of both data and model. To expand the
diversity of the logical samples, we propose an atom extension strategy
supported by equivalent logical formulas, to form new reasoning paths. From the
model perspective, we design a stack of transformer-style blocks. In
particular, we propose a path-attention module to joint model in-atom and
cross-atom relations with the high-order diffusion strategy. Experiments show
that PathReasoner achieves competitive performances on two logical reasoning
benchmarks and great generalization abilities. | cs.CL | PathReasoner: Modeling Reasoning Path with Equivalent Extension for Logical Question Answering |
2024-05-29T13:55:06Z | http://arxiv.org/pdf/2405.19094v1 | http://arxiv.org/abs/2405.19094v1 | Syrine Krichene, Francesco Piccinno, Fangyu Liu, Julian Martin Eisenschlos | 2024-05-29T13:55:06Z | cs.CL, cs.AI | Chart-to-summary generation can help explore data, communicate insights, and
help the visually impaired people. Multi-modal generative models have been used
to produce fluent summaries, but they can suffer from factual and perceptual
errors. In this work we present CHATS-CRITIC, a reference-free chart
summarization metric for scoring faithfulness. CHATS-CRITIC is composed of an
image-to-text model to recover the table from a chart, and a tabular entailment
model applied to score the summary sentence by sentence. We find that
CHATS-CRITIC evaluates the summary quality according to human ratings better
than reference-based metrics, either learned or n-gram based, and can be
further used to fix candidate summaries by removing not supported sentences. We
then introduce CHATS-PI, a chart-to-summary pipeline that leverages
CHATS-CRITIC during inference to fix and rank sampled candidates from any
chart-summarization model. We evaluate CHATS-PI and CHATS-CRITIC using human
raters, establishing state-of-the-art results on two popular chart-to-summary
datasets. | cs.CL | Faithful Chart Summarization with ChaTS-Pi |
2024-05-29T13:54:30Z | http://arxiv.org/pdf/2405.19093v1 | http://arxiv.org/abs/2405.19093v1 | Xindi Wang, Robert E. Mercer, Frank Rudzicz | 2024-05-29T13:54:30Z | cs.CL, cs.IR | The International Classification of Diseases (ICD) serves as a definitive
medical classification system encompassing a wide range of diseases and
conditions. The primary objective of ICD indexing is to allocate a subset of
ICD codes to a medical record, which facilitates standardized documentation and
management of various health conditions. Most existing approaches have suffered
from selecting the proper label subsets from an extremely large ICD collection
with a heavy long-tailed label distribution. In this paper, we leverage a
multi-stage ``retrieve and re-rank'' framework as a novel solution to ICD
indexing, via a hybrid discrete retrieval method, and re-rank retrieved
candidates with contrastive learning that allows the model to make more
accurate predictions from a simplified label space. The retrieval model is a
hybrid of auxiliary knowledge of the electronic health records (EHR) and a
discrete retrieval method (BM25), which efficiently collects high-quality
candidates. In the last stage, we propose a label co-occurrence guided
contrastive re-ranking model, which re-ranks the candidate labels by pulling
together the clinical notes with positive ICD codes. Experimental results show
the proposed method achieves state-of-the-art performance on a number of
measures on the MIMIC-III benchmark. | cs.CL | Multi-stage Retrieve and Re-rank Model for Automatic Medical Coding Recommendation |
2024-05-29T13:51:43Z | http://arxiv.org/pdf/2405.19088v1 | http://arxiv.org/abs/2405.19088v1 | Zhe Hu, Tuo Liang, Jing Li, Yiren Lu, Yunlai Zhou, Yiran Qiao, Jing Ma, Yu Yin | 2024-05-29T13:51:43Z | cs.CL, cs.CV | Recent advancements in large multimodal language models have demonstrated
remarkable proficiency across a wide range of tasks. Yet, these models still
struggle with understanding the nuances of human humor through juxtaposition,
particularly when it involves nonlinear narratives that underpin many jokes and
humor cues. This paper investigates this challenge by focusing on comics with
contradictory narratives, where each comic consists of two panels that create a
humorous contradiction. We introduce the YesBut benchmark, which comprises
tasks of varying difficulty aimed at assessing AI's capabilities in recognizing
and interpreting these comics, ranging from literal content comprehension to
deep narrative reasoning. Through extensive experimentation and analysis of
recent commercial or open-sourced large (vision) language models, we assess
their capability to comprehend the complex interplay of the narrative humor
inherent in these comics. Our results show that even state-of-the-art models
still lag behind human performance on this task. Our findings offer insights
into the current limitations and potential improvements for AI in understanding
human creative expressions. | cs.CL | Cracking the Code of Juxtaposition: Can AI Models Understand the Humorous Contradictions |
2024-05-29T13:49:44Z | http://arxiv.org/pdf/2405.19086v2 | http://arxiv.org/abs/2405.19086v2 | Renzhi Wang, Piji Li | 2024-06-02T02:32:31Z | cs.CL | Model editing aims to efficiently alter the behavior of Large Language Models
(LLMs) within a desired scope, while ensuring no adverse impact on other
inputs. Recent years have witnessed various model editing methods been
proposed. However, these methods either exhibit poor overall performance or
struggle to strike a balance between generalization and locality. We propose
MEMoE, a model editing adapter utilizing a Mixture of Experts (MoE)
architecture with a knowledge anchor routing strategy. MEMoE updates knowledge
using a bypass MoE structure, keeping the original parameters unchanged to
preserve the general ability of LLMs. And, the knowledge anchor routing ensures
that inputs requiring similar knowledge are routed to the same expert, thereby
enhancing the generalization of the updated knowledge. Experimental results
show the superiority of our approach over both batch editing and sequential
batch editing tasks, exhibiting exceptional overall performance alongside
outstanding balance between generalization and locality. Our code will be
available. | cs.CL | MEMoE: Enhancing Model Editing with Mixture of Experts Adaptors |
2024-05-29T13:44:07Z | http://arxiv.org/pdf/2405.19084v1 | http://arxiv.org/abs/2405.19084v1 | Xindi Wang, Robert E. Mercer, Frank Rudzicz | 2024-05-29T13:44:07Z | cs.CL | The International Classification of Diseases (ICD) is an authoritative
medical classification system of different diseases and conditions for clinical
and management purposes. ICD indexing assigns a subset of ICD codes to a
medical record. Since human coding is labour-intensive and error-prone, many
studies employ machine learning to automate the coding process. ICD coding is a
challenging task, as it needs to assign multiple codes to each medical document
from an extremely large hierarchically organized collection. In this paper, we
propose a novel approach for ICD indexing that adopts three ideas: (1) we use a
multi-level deep dilated residual convolution encoder to aggregate the
information from the clinical notes and learn document representations across
different lengths of the texts; (2) we formalize the task of ICD classification
with auxiliary knowledge of the medical records, which incorporates not only
the clinical texts but also different clinical code terminologies and drug
prescriptions for better inferring the ICD codes; and (3) we introduce a graph
convolutional network to leverage the co-occurrence patterns among ICD codes,
aiming to enhance the quality of label representations. Experimental results
show the proposed method achieves state-of-the-art performance on a number of
measures. | cs.CL | Auxiliary Knowledge-Induced Learning for Automatic Multi-Label Medical Document Classification |
2024-05-29T13:34:32Z | http://arxiv.org/pdf/2405.19076v2 | http://arxiv.org/abs/2405.19076v2 | Markus J. Buehler | 2024-06-02T15:03:24Z | cs.CV, cond-mat.mes-hall, cond-mat.mtrl-sci, cs.CL, cs.LG | We present Cephalo, a series of multimodal vision large language models
(V-LLMs) designed for materials science applications, integrating visual and
linguistic data for enhanced understanding and interaction within human-AI and
multi-agent AI frameworks. A key innovation of Cephalo is its advanced dataset
generation method, which employs a sophisticated algorithm to accurately detect
and separate images and their corresponding textual descriptions from PDF
documents, such as scientific papers. The method includes a careful refinement
of image-text pairs through integrated vision and language processing, ensuring
high-quality, contextually relevant, and well reasoned training data. Cephalo
is trained on integrated image and text data extracted from thousands of
scientific papers and science-focused Wikipedia pages demonstrates can
interpret complex visual scenes, generate precise language descriptions, and
answer queries about images effectively. The combination of a vision encoder
with an autoregressive transformer supports complex natural language
understanding in an integrated model, which can be coupled with other
generative methods to create an image-to-text-to-image or image-to-text-to-3D
pipeline. To explore the development of larger models from smaller ones, we
report both mixture-of-expert methods and model merging. These hybrid
approaches allow us to leverage the domain-specific expertise and general
conversational capabilities to harness the strengths of multiple models. We
examine the models in diverse use cases that incorporate biological materials,
fracture and engineering analysis, protein biophysics, and bio-inspired design
based on insect behavior. Generative applications include bio-inspired designs,
including pollen-inspired architected materials, as well as the synthesis of
bio-inspired material microstructures from a photograph of a solar eclipse. | cs.CV | Cephalo: Multi-Modal Vision-Language Models for Bio-Inspired Materials Analysis and Design |
2024-05-29T12:32:08Z | http://arxiv.org/pdf/2405.19041v1 | http://arxiv.org/abs/2405.19041v1 | Chen Wang, Minpeng Liao, Zhongqiang Huang, Jiajun Zhang | 2024-05-29T12:32:08Z | cs.CL, cs.SD, eess.AS | Recent end-to-end approaches have shown promise in extending large language
models (LLMs) to speech inputs, but face limitations in directly assessing and
optimizing alignment quality and fail to achieve fine-grained alignment due to
speech-text length mismatch. We introduce BLSP-KD, a novel approach for
Bootstrapping Language-Speech Pretraining via Knowledge Distillation, which
addresses these limitations through two key techniques. First, it optimizes
speech-text alignment by minimizing the divergence between the LLM's next-token
prediction distributions for speech and text inputs using knowledge
distillation. Second, it employs a continuous-integrate-andfire strategy to
segment speech into tokens that correspond one-to-one with text tokens,
enabling fine-grained alignment. We also introduce Partial LoRA (PLoRA), a new
adaptation method supporting LLM finetuning for speech inputs under knowledge
distillation. Quantitative evaluation shows that BLSP-KD outperforms previous
end-to-end baselines and cascaded systems with comparable scale of parameters,
facilitating general instruction-following capabilities for LLMs with speech
inputs. This approach provides new possibilities for extending LLMs to spoken
language interactions. | cs.CL | BLSP-KD: Bootstrapping Language-Speech Pre-training via Knowledge Distillation |
2024-05-29T12:12:09Z | http://arxiv.org/pdf/2405.19026v1 | http://arxiv.org/abs/2405.19026v1 | Andrew Zhao, Quentin Xu, Matthieu Lin, Shenzhi Wang, Yong-jin Liu, Zilong Zheng, Gao Huang | 2024-05-29T12:12:09Z | cs.LG, cs.AI, cs.CL, cs.CR | Recent advances in large language models (LLMs) have made them indispensable,
raising significant concerns over managing their safety. Automated red teaming
offers a promising alternative to the labor-intensive and error-prone manual
probing for vulnerabilities, providing more consistent and scalable safety
evaluations. However, existing approaches often compromise diversity by
focusing on maximizing attack success rate. Additionally, methods that decrease
the cosine similarity from historical embeddings with semantic diversity
rewards lead to novelty stagnation as history grows. To address these issues,
we introduce DiveR-CT, which relaxes conventional constraints on the objective
and semantic reward, granting greater freedom for the policy to enhance
diversity. Our experiments demonstrate DiveR-CT's marked superiority over
baselines by 1) generating data that perform better in various diversity
metrics across different attack success rate levels, 2) better-enhancing
resiliency in blue team models through safety tuning based on collected data,
3) allowing dynamic control of objective weights for reliable and controllable
attack success rates, and 4) reducing susceptibility to reward
overoptimization. Project details and code can be found at
https://andrewzh112.github.io/#diverct. | cs.LG | DiveR-CT: Diversity-enhanced Red Teaming with Relaxing Constraints |
2024-05-29T11:48:27Z | http://arxiv.org/pdf/2405.19010v1 | http://arxiv.org/abs/2405.19010v1 | Hao Zhang, Yuyang Zhang, Xiaoguang Li, Wenxuan Shi, Haonan Xu, Huanshuo Liu, Yasheng Wang, Lifeng Shang, Qun Liu, Yong Liu, Ruiming Tang | 2024-05-29T11:48:27Z | cs.CL, cs.AI, cs.IR | Integrating external knowledge into large language models (LLMs) presents a
promising solution to overcome the limitations imposed by their antiquated and
static parametric memory. Prior studies, however, have tended to over-reliance
on external knowledge, underestimating the valuable contributions of an LLMs'
intrinsic parametric knowledge. The efficacy of LLMs in blending external and
parametric knowledge remains largely unexplored, especially in cases where
external knowledge is incomplete and necessitates supplementation by their
parametric knowledge. We propose to deconstruct knowledge fusion into four
distinct scenarios, offering the first thorough investigation of LLM behavior
across each. We develop a systematic pipeline for data construction and
knowledge infusion to simulate these fusion scenarios, facilitating a series of
controlled experiments. Our investigation reveals that enhancing parametric
knowledge within LLMs can significantly bolster their capability for knowledge
integration. Nonetheless, we identify persistent challenges in memorizing and
eliciting parametric knowledge, and determining parametric knowledge
boundaries. Our findings aim to steer future explorations on harmonizing
external and parametric knowledge within LLMs. | cs.CL | Evaluating the External and Parametric Knowledge Fusion of Large Language Models |
2024-05-29T11:11:07Z | http://arxiv.org/pdf/2405.18991v1 | http://arxiv.org/abs/2405.18991v1 | Jiaqi Xu, Xinyi Zou, Kunzhe Huang, Yunkuo Chen, Bo Liu, MengLi Cheng, Xing Shi, Jun Huang | 2024-05-29T11:11:07Z | cs.CV, cs.CL, cs.MM | This paper presents EasyAnimate, an advanced method for video generation that
leverages the power of transformer architecture for high-performance outcomes.
We have expanded the DiT framework originally designed for 2D image synthesis
to accommodate the complexities of 3D video generation by incorporating a
motion module block. It is used to capture temporal dynamics, thereby ensuring
the production of consistent frames and seamless motion transitions. The motion
module can be adapted to various DiT baseline methods to generate video with
different styles. It can also generate videos with different frame rates and
resolutions during both training and inference phases, suitable for both images
and videos. Moreover, we introduce slice VAE, a novel approach to condense the
temporal axis, facilitating the generation of long duration videos. Currently,
EasyAnimate exhibits the proficiency to generate videos with 144 frames. We
provide a holistic ecosystem for video production based on DiT, encompassing
aspects such as data pre-processing, VAE training, DiT models training (both
the baseline model and LoRA model), and end-to-end video inference. Code is
available at: https://github.com/aigc-apps/EasyAnimate. We are continuously
working to enhance the performance of our method. | cs.CV | EasyAnimate: A High-Performance Long Video Generation Method based on Transformer Architecture |
2024-05-29T10:37:28Z | http://arxiv.org/pdf/2405.18974v1 | http://arxiv.org/abs/2405.18974v1 | Songtao Liu, Bang Wang, Wei Xiang, Han Xu, Minghua Xu | 2024-05-29T10:37:28Z | cs.CL | Multifaceted ideology detection (MID) aims to detect the ideological leanings
of texts towards multiple facets. Previous studies on ideology detection mainly
focus on one generic facet and ignore label semantics and explanatory
descriptions of ideologies, which are a kind of instructive information and
reveal the specific concepts of ideologies. In this paper, we develop a novel
concept semantics-enhanced framework for the MID task. Specifically, we propose
a bidirectional iterative concept flow (BICo) method to encode multifaceted
ideologies. BICo enables the concepts to flow across levels of the schema tree
and enriches concept representations with multi-granularity semantics.
Furthermore, we explore concept attentive matching and concept-guided
contrastive learning strategies to guide the model to capture ideology features
with the learned concept semantics. Extensive experiments on the benchmark
dataset show that our approach achieves state-of-the-art performance in MID,
including in the cross-topic scenario. | cs.CL | Encoding Hierarchical Schema via Concept Flow for Multifaceted Ideology Detection |
2024-05-29T10:08:31Z | http://arxiv.org/pdf/2405.18952v2 | http://arxiv.org/abs/2405.18952v2 | Peter Devine | 2024-06-01T02:18:06Z | cs.CL, cs.AI, cs.LG | Training Large Language Models (LLMs) with Reinforcement Learning from AI
Feedback (RLAIF) aligns model outputs more closely with human preferences. This
involves an evaluator model ranking multiple candidate responses to user
prompts. However, the rankings from popular evaluator models such as GPT-4 can
be inconsistent. We propose the Repeat Ranking method - where we evaluate the
same responses multiple times and train only on those responses which are
consistently ranked. Using 2,714 prompts in 62 languages, we generated
responses from 7 top multilingual LLMs and had GPT-4 rank them five times each.
Evaluating on MT-Bench chat benchmarks in six languages, our method
outperformed the standard practice of training on all available prompts. Our
work highlights the quality versus quantity trade-off in RLAIF dataset
generation and offers a stackable strategy for enhancing dataset and thus model
quality. | cs.CL | Are You Sure? Rank Them Again: Repeated Ranking For Better Preference Datasets |
2024-05-29T09:43:48Z | http://arxiv.org/pdf/2405.18937v1 | http://arxiv.org/abs/2405.18937v1 | Junjie Fei, Mahmoud Ahmed, Jian Ding, Eslam Mohamed Bakr, Mohamed Elhoseiny | 2024-05-29T09:43:48Z | cs.CV, cs.CL | While 3D MLLMs have achieved significant progress, they are restricted to
object and scene understanding and struggle to understand 3D spatial structures
at the part level. In this paper, we introduce Kestrel, representing a novel
approach that empowers 3D MLLMs with part-aware understanding, enabling better
interpretation and segmentation grounding of 3D objects at the part level.
Despite its significance, the current landscape lacks tasks and datasets that
endow and assess this capability. Therefore, we propose two novel tasks: (1)
Part-Aware Point Grounding, the model is tasked with directly predicting a
part-level segmentation mask based on user instructions, and (2) Part-Aware
Point Grounded Captioning, the model provides a detailed caption that includes
part-level descriptions and their corresponding masks. To support learning and
evaluating for these tasks, we introduce 3DCoMPaT Grounded Instructions Dataset
(3DCoMPaT-GRIN). 3DCoMPaT-GRIN Vanilla, comprising 789k part-aware point
cloud-instruction-segmentation mask triplets, is used to evaluate MLLMs'
ability of part-aware segmentation grounding. 3DCoMPaT-GRIN Grounded Caption,
containing 107k part-aware point cloud-instruction-grounded caption triplets,
assesses both MLLMs' part-aware language comprehension and segmentation
grounding capabilities. Our introduced tasks, dataset, and Kestrel represent a
preliminary effort to bridge the gap between human cognition and 3D MLLMs,
i.e., the ability to perceive and engage with the environment at both global
and part levels. Extensive experiments on the 3DCoMPaT-GRIN show that Kestrel
can generate user-specified segmentation masks, a capability not present in any
existing 3D MLLM. Kestrel thus established a benchmark for evaluating the
part-aware language comprehension and segmentation grounding of 3D objects.
Project page at https://feielysia.github.io/Kestrel.github.io/ | cs.CV | Kestrel: Point Grounding Multimodal LLM for Part-Aware 3D Vision-Language Understanding |
2024-05-29T09:25:49Z | http://arxiv.org/pdf/2405.18922v1 | http://arxiv.org/abs/2405.18922v1 | Chenze Shao, Fandong Meng, Jiali Zeng, Jie Zhou | 2024-05-29T09:25:49Z | cs.CL | Neural Machine Translation (NMT) has made remarkable progress over the past
years. However, under-translation and over-translation remain two challenging
problems in state-of-the-art NMT systems. In this work, we conduct an in-depth
analysis on the underlying cause of under-translation in NMT, providing an
explanation from the perspective of decoding objective. To optimize the beam
search objective, the model tends to overlook words it is less confident about,
leading to the under-translation phenomenon. Correspondingly, the model's
confidence in predicting the End Of Sentence (EOS) diminishes when
under-translation occurs, serving as a mild penalty for under-translated
candidates. Building upon this analysis, we propose employing the confidence of
predicting EOS as a detector for under-translation, and strengthening the
confidence-based penalty to penalize candidates with a high risk of
under-translation. Experiments on both synthetic and real-world data show that
our method can accurately detect and rectify under-translated outputs, with
minor impact on other correct translations. | cs.CL | Understanding and Addressing the Under-Translation Problem from the Perspective of Decoding Objective |
2024-05-29T09:17:46Z | http://arxiv.org/pdf/2405.18915v1 | http://arxiv.org/abs/2405.18915v1 | Jiachun Li, Pengfei Cao, Yubo Chen, Kang Liu, Jun Zhao | 2024-05-29T09:17:46Z | cs.CL, cs.AI | Large language models (LLMs) suffer from serious unfaithful chain-of-thought
(CoT) issues. Previous work attempts to measure and explain it but lacks
in-depth analysis within CoTs and does not consider the interactions among all
reasoning components jointly. In this paper, we first study the CoT
faithfulness issue at the granularity of CoT steps, identify two reasoning
paradigms: centralized reasoning and distributed reasoning, and find their
relationship with faithfulness. Subsequently, we conduct a joint analysis of
the causal relevance among the context, CoT, and answer during reasoning. The
result proves that, when the LLM predicts answers, it can recall correct
information missing in the CoT from the context, leading to unfaithfulness
issues. Finally, we propose the inferential bridging method to mitigate this
issue, in which we use the attribution method to recall information as hints
for CoT generation and filter out noisy CoTs based on their semantic
consistency and attribution scores. Extensive experiments demonstrate that our
approach effectively alleviates the unfaithful CoT problem. | cs.CL | Towards Faithful Chain-of-Thought: Large Language Models are Bridging Reasoners |
2024-05-29T09:09:00Z | http://arxiv.org/pdf/2405.18906v1 | http://arxiv.org/abs/2405.18906v1 | Chenze Shao, Fandong Meng, Yijin Liu, Jie Zhou | 2024-05-29T09:09:00Z | cs.CL, cs.LG | Language generation based on maximum likelihood estimation (MLE) has become
the fundamental approach for text generation. Maximum likelihood estimation is
typically performed by minimizing the log-likelihood loss, also known as the
logarithmic score in statistical decision theory. The logarithmic score is
strictly proper in the sense that it encourages honest forecasts, where the
expected score is maximized only when the model reports true probabilities.
Although many strictly proper scoring rules exist, the logarithmic score is the
only local scoring rule among them that depends exclusively on the probability
of the observed sample, making it capable of handling the exponentially large
sample space of natural text. In this work, we propose a straightforward
strategy for adapting scoring rules to language generation, allowing for
language modeling with any non-local scoring rules. Leveraging this strategy,
we train language generation models using two classic strictly proper scoring
rules, the Brier score and the Spherical score, as alternatives to the
logarithmic score. Experimental results indicate that simply substituting the
loss function, without adjusting other hyperparameters, can yield substantial
improvements in model's generation capabilities. Moreover, these improvements
can scale up to large language models (LLMs) such as LLaMA-7B and LLaMA-13B.
Source code: \url{https://github.com/shaochenze/ScoringRulesLM}. | cs.CL | Language Generation with Strictly Proper Scoring Rules |
2024-05-29T09:06:18Z | http://arxiv.org/pdf/2406.16899v1 | http://arxiv.org/abs/2406.16899v1 | Yuni Susanti, Nina Holsmoelle | 2024-05-29T09:06:18Z | cs.CL, cs.AI | This work aims toward an application of natural language processing (NLP)
technology for automatic verification of causal graphs using text sources. A
causal graph is often derived from unsupervised causal discovery methods and
requires manual evaluation from human experts. NLP technologies, i.e., Large
Language Models (LLMs) such as BERT and ChatGPT, can potentially be used to
verify the resulted causal graph by predicting if causal relation can be
observed between node pairs based on the textual context. In this work, we
compare the performance of two types of NLP models: (1) Pre-trained language
models fine-tuned for causal relation classification task and, (2) prompt-based
LLMs. Contrasted to previous studies where prompt-based LLMs work relatively
well over a set of diverse tasks, preliminary experiments on biomedical and
open-domain datasets suggest that the fine-tuned models far outperform the
prompt-based LLMs, up to 20.5 points improvement of F1 score. We shared the
code and the pre-processed datasets in our repository. | cs.CL | Prompt-based vs. Fine-tuned LLMs Toward Causal Graph Verification |
2024-05-29T08:32:37Z | http://arxiv.org/pdf/2405.18874v1 | http://arxiv.org/abs/2405.18874v1 | Riccardo Rende, Luciano Loris Viteritti | 2024-05-29T08:32:37Z | cond-mat.dis-nn, cs.CL, physics.comp-ph | The dot product attention mechanism, originally designed for natural language
processing (NLP) tasks, is a cornerstone of modern Transformers. It adeptly
captures semantic relationships between word pairs in sentences by computing a
similarity overlap between queries and keys. In this work, we explore the
suitability of Transformers, focusing on their attention mechanisms, in the
specific domain of the parametrization of variational wave functions to
approximate ground states of quantum many-body spin Hamiltonians. Specifically,
we perform numerical simulations on the two-dimensional $J_1$-$J_2$ Heisenberg
model, a common benchmark in the field of quantum-many body systems on lattice.
By comparing the performance of standard attention mechanisms with a simplified
version that excludes queries and keys, relying solely on positions, we achieve
competitive results while reducing computational cost and parameter usage.
Furthermore, through the analysis of the attention maps generated by standard
attention mechanisms, we show that the attention weights become effectively
input-independent at the end of the optimization. We support the numerical
results with analytical calculations, providing physical insights of why
queries and keys should be, in principle, omitted from the attention mechanism
when studying large systems. Interestingly, the same arguments can be extended
to the NLP domain, in the limit of long input sentences. | cond-mat.dis-nn | Are queries and keys always relevant? A case study on Transformer wave functions |
2024-05-29T08:31:16Z | http://arxiv.org/pdf/2405.18870v2 | http://arxiv.org/abs/2405.18870v2 | Winnie Street, John Oliver Siy, Geoff Keeling, Adrien Baranes, Benjamin Barnett, Michael McKibben, Tatenda Kanyere, Alison Lentz, Blaise Aguera y Arcas, Robin I. M. Dunbar | 2024-05-31T12:45:50Z | cs.AI, cs.CL, cs.HC, I.2.7; H.1.2 | This paper examines the extent to which large language models (LLMs) have
developed higher-order theory of mind (ToM); the human ability to reason about
multiple mental and emotional states in a recursive manner (e.g. I think that
you believe that she knows). This paper builds on prior work by introducing a
handwritten test suite -- Multi-Order Theory of Mind Q&A -- and using it to
compare the performance of five LLMs to a newly gathered adult human benchmark.
We find that GPT-4 and Flan-PaLM reach adult-level and near adult-level
performance on ToM tasks overall, and that GPT-4 exceeds adult performance on
6th order inferences. Our results suggest that there is an interplay between
model size and finetuning for the realisation of ToM abilities, and that the
best-performing LLMs have developed a generalised capacity for ToM. Given the
role that higher-order ToM plays in a wide range of cooperative and competitive
human behaviours, these findings have significant implications for user-facing
LLM applications. | cs.AI | LLMs achieve adult human performance on higher-order theory of mind tasks |
2024-05-29T07:56:08Z | http://arxiv.org/pdf/2405.18845v1 | http://arxiv.org/abs/2405.18845v1 | Silvia García Méndez, Fátima Leal, Benedita Malheiro, Juan Carlos Burguillo Rial, Bruno Veloso, Adriana E. Chis, Horacio González Vélez | 2024-05-29T07:56:08Z | cs.CL, cs.AI, cs.LG | Data crowdsourcing is a data acquisition process where groups of voluntary
contributors feed platforms with highly relevant data ranging from news,
comments, and media to knowledge and classifications. It typically processes
user-generated data streams to provide and refine popular services such as
wikis, collaborative maps, e-commerce sites, and social networks. Nevertheless,
this modus operandi raises severe concerns regarding ill-intentioned data
manipulation in adversarial environments. This paper presents a simulation,
modelling, and classification approach to automatically identify human and
non-human (bots) as well as benign and malign contributors by using data
fabrication to balance classes within experimental data sets, data stream
modelling to build and update contributor profiles and, finally, autonomic data
stream classification. By employing WikiVoyage - a free worldwide wiki travel
guide open to contribution from the general public - as a testbed, our approach
proves to significantly boost the confidence and quality of the classifier by
using a class-balanced data stream, comprising both real and synthetic data.
Our empirical results show that the proposed method distinguishes between
benign and malign bots as well as human contributors with a classification
accuracy of up to 92 %. | cs.CL | Simulation, Modelling and Classification of Wiki Contributors: Spotting The Good, The Bad, and The Ugly |
2024-05-29T07:03:31Z | http://arxiv.org/pdf/2405.18822v1 | http://arxiv.org/abs/2405.18822v1 | Zhanhao Hu, Julien Piet, Geng Zhao, Jiantao Jiao, David Wagner | 2024-05-29T07:03:31Z | cs.CL | Current LLMs are generally aligned to follow safety requirements and tend to
refuse toxic prompts. However, LLMs can fail to refuse toxic prompts or be
overcautious and refuse benign examples. In addition, state-of-the-art toxicity
detectors have low TPRs at low FPR, incurring high costs in real-world
applications where toxic examples are rare. In this paper, we explore
Moderation Using LLM Introspection (MULI), which detects toxic prompts using
the information extracted directly from LLMs themselves. We found significant
gaps between benign and toxic prompts in the distribution of alternative
refusal responses and in the distribution of the first response token's logits.
These gaps can be used to detect toxicities: We show that a toy model based on
the logits of specific starting tokens gets reliable performance, while
requiring no training or additional computational cost. We build a more robust
detector using a sparse logistic regression model on the first response token
logits, which greatly exceeds SOTA detectors under multiple metrics. | cs.CL | Toxicity Detection for Free |
2024-05-29T05:32:50Z | http://arxiv.org/pdf/2405.18776v1 | http://arxiv.org/abs/2405.18776v1 | Qin Yang, Meisam Mohammad, Han Wang, Ali Payani, Ashish Kundu, Kai Shu, Yan Yan, Yuan Hong | 2024-05-29T05:32:50Z | cs.CR, cs.CL, cs.LG | Differentially Private Stochastic Gradient Descent (DP-SGD) and its variants
have been proposed to ensure rigorous privacy for fine-tuning large-scale
pre-trained language models. However, they rely heavily on the Gaussian
mechanism, which may overly perturb the gradients and degrade the accuracy,
especially in stronger privacy regimes (e.g., the privacy budget $\epsilon <
3$). To address such limitations, we propose a novel Language Model-based
Optimal Differential Privacy (LMO-DP) mechanism, which takes the first step to
enable the tight composition of accurately fine-tuning (large) language models
with a sub-optimal DP mechanism, even in strong privacy regimes (e.g., $0.1\leq
\epsilon<3$). Furthermore, we propose a novel offline optimal noise search
method to efficiently derive the sub-optimal DP that significantly reduces the
noise magnitude. For instance, fine-tuning RoBERTa-large (with 300M parameters)
on the SST-2 dataset can achieve an accuracy of 92.20% (given $\epsilon=0.3$,
$\delta=10^{-10}$) by drastically outperforming the Gaussian mechanism (e.g.,
$\sim 50\%$ for small $\epsilon$ and $\delta$). We also draw similar findings
on the text generation tasks on GPT-2. Finally, to our best knowledge, LMO-DP
is also the first solution to accurately fine-tune Llama-2 with strong
differential privacy guarantees. The code will be released soon and available
upon request. | cs.CR | LMO-DP: Optimizing the Randomization Mechanism for Differentially Private Fine-Tuning (Large) Language Models |
2024-05-29T04:04:36Z | http://arxiv.org/pdf/2405.18742v1 | http://arxiv.org/abs/2405.18742v1 | Reed Perkins, Dan Ventura | 2024-05-29T04:04:36Z | cs.AI, cs.CL | We outline a solution to the challenge of musical phrase segmentation that
uses grammatical induction algorithms, a class of algorithms which infer a
context-free grammar from an input sequence. We analyze the performance of five
grammatical induction algorithms on three datasets using various musical
viewpoint combinations. Our experiments show that the LONGESTFIRST algorithm
achieves the best F1 scores across all three datasets and that input encodings
that include the duration viewpoint result in the best performance. | cs.AI | Musical Phrase Segmentation via Grammatical Induction |
2024-05-29T04:04:05Z | http://arxiv.org/pdf/2405.18741v2 | http://arxiv.org/abs/2405.18741v2 | Xiao Peng, Tao Liu, Ying Wang | 2024-06-03T08:35:07Z | cs.CL, cs.AI | Large language models (LLMs) like ChatGPT, Gemini, or LLaMA have been
trending recently, demonstrating considerable advancement and generalizability
power in countless domains. However, LLMs create an even bigger black box
exacerbating opacity, with interpretability limited to few approaches. The
uncertainty and opacity embedded in LLMs' nature restrict their application in
high-stakes domains like financial fraud, phishing, etc. Current approaches
mainly rely on traditional textual classification with posterior interpretable
algorithms, suffering from attackers who may create versatile adversarial
samples to break the system's defense, forcing users to make trade-offs between
efficiency and robustness. To address this issue, we propose a novel cascading
framework called Genshin (General Shield for Natural Language Processing with
Large Language Models), utilizing LLMs as defensive one-time plug-ins. Unlike
most applications of LLMs that try to transform text into something new or
structural, Genshin uses LLMs to recover text to its original state. Genshin
aims to combine the generalizability of the LLM, the discrimination of the
median model, and the interpretability of the simple model. Our experiments on
the task of sentimental analysis and spam detection have shown fatal flaws of
the current median models and exhilarating results on LLMs' recovery ability,
demonstrating that Genshin is both effective and efficient. In our ablation
study, we unearth several intriguing observations. Utilizing the LLM defender,
a tool derived from the 4th paradigm, we have reproduced BERT's 15% optimal
mask rate results in the 3rd paradigm of NLP. Additionally, when employing the
LLM as a potential adversarial tool, attackers are capable of executing
effective attacks that are nearly semantically lossless. | cs.CL | Genshin: General Shield for Natural Language Processing with Large Language Models |
2024-05-29T04:00:41Z | http://arxiv.org/pdf/2405.18740v1 | http://arxiv.org/abs/2405.18740v1 | Jialiang Xu, Michael Moor, Jure Leskovec | 2024-05-29T04:00:41Z | cs.CL | Despite impressive advances in recent multimodal large language models
(MLLMs), state-of-the-art models such as from the GPT-4 suite still struggle
with knowledge-intensive tasks. To address this, we consider Reverse Image
Retrieval (RIR) augmented generation, a simple yet effective strategy to
augment MLLMs with web-scale reverse image search results. RIR robustly
improves knowledge-intensive visual question answering (VQA) of GPT-4V by
37-43%, GPT-4 Turbo by 25-27%, and GPT-4o by 18-20% in terms of open-ended VQA
evaluation metrics. To our surprise, we discover that RIR helps the model to
better access its own world knowledge. Concretely, our experiments suggest that
RIR augmentation helps by providing further visual and textual cues without
necessarily containing the direct answer to a query. In addition, we elucidate
cases in which RIR can hurt performance and conduct a human evaluation.
Finally, we find that the overall advantage of using RIR makes it difficult for
an agent that can choose to use RIR to perform better than an approach where
RIR is the default setting. | cs.CL | Reverse Image Retrieval Cues Parametric Memory in Multimodal LLMs |
2024-05-29T03:17:16Z | http://arxiv.org/pdf/2405.18727v1 | http://arxiv.org/abs/2405.18727v1 | Huanshuo Liu, Hao Zhang, Zhijiang Guo, Kuicai Dong, Xiangyang Li, Yi Quan Lee, Cong Zhang, Yong Liu | 2024-05-29T03:17:16Z | cs.CL, cs.AI, cs.IR | Retrieval-augmented generation (RAG) has emerged as a promising solution for
mitigating hallucinations of large language models (LLMs) with retrieved
external knowledge. Adaptive RAG enhances this approach by dynamically
assessing the retrieval necessity, aiming to balance external and internal
knowledge usage. However, existing adaptive RAG methods primarily realize
retrieval on demand by relying on superficially verbalize-based or
probability-based feedback of LLMs, or directly fine-tuning LLMs via carefully
crafted datasets, resulting in unreliable retrieval necessity decisions, heavy
extra costs, and sub-optimal response generation. We present the first attempts
to delve into the internal states of LLMs to mitigate such issues by
introducing an effective probe-guided adaptive RAG framework, termed CtrlA.
Specifically, CtrlA employs an honesty probe to regulate the LLM's behavior by
manipulating its representations for increased honesty, and a confidence probe
to monitor the internal states of LLM and assess confidence levels, determining
the retrieval necessity during generation. Experiments show that CtrlA is
superior to existing adaptive RAG methods on a diverse set of tasks, the
honesty control can effectively make LLMs more honest and confidence monitoring
is proven to be a promising indicator of retrieval trigger. Our codes are
available at https://github.com/HSLiu-Initial/CtrlA.git. | cs.CL | CtrlA: Adaptive Retrieval-Augmented Generation via Probe-Guided Control |
2024-05-29T03:05:59Z | http://arxiv.org/pdf/2405.18721v2 | http://arxiv.org/abs/2405.18721v2 | Bingqian Lin, Yunshuang Nie, Ziming Wei, Yi Zhu, Hang Xu, Shikui Ma, Jianzhuang Liu, Xiaodan Liang | 2024-06-05T09:59:21Z | cs.CV, cs.AI, cs.CL | Vision-Language Navigation (VLN) requires the agent to follow language
instructions to reach a target position. A key factor for successful navigation
is to align the landmarks implied in the instruction with diverse visual
observations. However, previous VLN agents fail to perform accurate modality
alignment especially in unexplored scenes, since they learn from limited
navigation data and lack sufficient open-world alignment knowledge. In this
work, we propose a new VLN paradigm, called COrrectable LaNdmark DiScOvery via
Large ModEls (CONSOLE). In CONSOLE, we cast VLN as an open-world sequential
landmark discovery problem, by introducing a novel correctable landmark
discovery scheme based on two large models ChatGPT and CLIP. Specifically, we
use ChatGPT to provide rich open-world landmark cooccurrence commonsense, and
conduct CLIP-driven landmark discovery based on these commonsense priors. To
mitigate the noise in the priors due to the lack of visual constraints, we
introduce a learnable cooccurrence scoring module, which corrects the
importance of each cooccurrence according to actual observations for accurate
landmark discovery. We further design an observation enhancement strategy for
an elegant combination of our framework with different VLN agents, where we
utilize the corrected landmark features to obtain enhanced observation features
for action decision. Extensive experimental results on multiple popular VLN
benchmarks (R2R, REVERIE, R4R, RxR) show the significant superiority of CONSOLE
over strong baselines. Especially, our CONSOLE establishes the new
state-of-the-art results on R2R and R4R in unseen scenarios. Code is available
at https://github.com/expectorlin/CONSOLE. | cs.CV | Correctable Landmark Discovery via Large Models for Vision-Language Navigation |
2024-05-29T02:57:15Z | http://arxiv.org/pdf/2405.18719v2 | http://arxiv.org/abs/2405.18719v2 | Olga Golovneva, Tianlu Wang, Jason Weston, Sainbayar Sukhbaatar | 2024-05-30T17:51:53Z | cs.CL, cs.AI | The attention mechanism is a critical component of Large Language Models
(LLMs) that allows tokens in a sequence to interact with each other, but is
order-invariant. Incorporating position encoding (PE) makes it possible to
address by position, such as attending to the i-th token. However, current PE
methods use token counts to derive position, and thus cannot generalize to
higher levels of abstraction, such as attending to the i-th sentence. In this
paper, we propose a new position encoding method, Contextual Position Encoding
(CoPE), that allows positions to be conditioned on context by incrementing
position only on certain tokens determined by the model. This allows more
general position addressing such as attending to the $i$-th particular word,
noun, or sentence. We show that CoPE can solve the selective copy, counting and
Flip-Flop tasks where popular position embeddings fail, and improves perplexity
on language modeling and coding tasks. | cs.CL | Contextual Position Encoding: Learning to Count What's Important |
2024-05-29T02:57:07Z | http://arxiv.org/pdf/2405.18718v1 | http://arxiv.org/abs/2405.18718v1 | Fengshuo Bai, Mingzhi Wang, Zhaowei Zhang, Boyuan Chen, Yinda Xu, Ying Wen, Yaodong Yang | 2024-05-29T02:57:07Z | cs.CL | With recent advancements in large language models (LLMs), alignment has
emerged as an effective technique for keeping LLMs consensus with human intent.
Current methods primarily involve direct training through Supervised
Fine-tuning (SFT) or Reinforcement Learning from Human Feedback (RLHF), both of
which require substantial computational resources and extensive ground truth
data. This paper explores an efficient method for aligning black-box large
models using smaller models, introducing a model-agnostic and lightweight
Bayesian Persuasion Alignment framework. We formalize this problem as an
optimization of the signaling strategy from the small model's perspective. In
the persuasion process, the small model (Advisor) observes the information item
(i.e., state) and persuades large models (Receiver) to elicit improved
responses. The Receiver then generates a response based on the input, the
signal from the Advisor, and its updated belief about the information item.
Through training using our framework, we demonstrate that the Advisor can
significantly enhance the performance of various Receivers across a range of
tasks. We theoretically analyze our persuasion framework and provide an upper
bound on the Advisor's regret, confirming its effectiveness in learning the
optimal signaling strategy. Our Empirical results demonstrates that GPT-2 can
significantly improve the performance of various models, achieving an average
enhancement of 16.1% in mathematical reasoning ability and 13.7% in code
generation. We hope our work can provide an initial step toward rethinking the
alignment framework from the Bayesian Persuasion perspective. | cs.CL | Efficient Model-agnostic Alignment via Bayesian Persuasion |
2024-05-29T02:44:12Z | http://arxiv.org/pdf/2405.18711v1 | http://arxiv.org/abs/2405.18711v1 | Zhihui Xie, Jizhou Guo, Tong Yu, Shuai Li | 2024-05-29T02:44:12Z | cs.AI, cs.CL | Large language models (LLMs) have demonstrated impressive capabilities in
various reasoning tasks, aided by techniques like chain-of-thought (CoT)
prompting that elicits verbalized reasoning. However, LLMs often generate text
with obvious mistakes and contradictions, raising doubts about their ability to
robustly process and utilize generated rationales. In this work, we investigate
CoT reasoning in LLMs through the lens of internal representations, focusing on
how these representations are influenced by generated rationales. Our
preliminary analysis reveals that while generated rationales improve answer
accuracy, inconsistencies emerge between the model's internal representations
in middle layers and those in final layers, potentially undermining the
reliability of their reasoning processes. To address this, we propose internal
consistency as a measure of the model's confidence by examining the agreement
of latent predictions decoded from intermediate layers. Extensive empirical
studies across different models and datasets demonstrate that internal
consistency effectively distinguishes between correct and incorrect reasoning
paths. Motivated by this, we propose a new approach to calibrate CoT reasoning
by up-weighting reasoning paths with high internal consistency, resulting in a
significant boost in reasoning performance. Further analysis uncovers distinct
patterns in attention and feed-forward modules across layers, providing
insights into the emergence of internal inconsistency. In summary, our results
demonstrate the potential of using internal representations for self-evaluation
of LLMs. | cs.AI | Calibrating Reasoning in Language Models with Internal Consistency |
2024-05-29T01:49:20Z | http://arxiv.org/pdf/2405.18688v1 | http://arxiv.org/abs/2405.18688v1 | Fengshuo Bai, Rui Zhao, Hongming Zhang, Sijia Cui, Ying Wen, Yaodong Yang, Bo Xu, Lei Han | 2024-05-29T01:49:20Z | cs.LG, cs.AI, cs.CL | Preference-based reinforcement learning (PbRL) has shown impressive
capabilities in training agents without reward engineering. However, a notable
limitation of PbRL is its dependency on substantial human feedback. This
dependency stems from the learning loop, which entails accurate reward learning
compounded with value/policy learning, necessitating a considerable number of
samples. To boost the learning loop, we propose SEER, an efficient PbRL method
that integrates label smoothing and policy regularization techniques. Label
smoothing reduces overfitting of the reward model by smoothing human preference
labels. Additionally, we bootstrap a conservative estimate $\widehat{Q}$ using
well-supported state-action pairs from the current replay memory to mitigate
overestimation bias and utilize it for policy learning regularization. Our
experimental results across a variety of complex tasks, both in online and
offline settings, demonstrate that our approach improves feedback efficiency,
outperforming state-of-the-art methods by a large margin. Ablation studies
further reveal that SEER achieves a more accurate Q-function compared to prior
work. | cs.LG | Efficient Preference-based Reinforcement Learning via Aligned Experience Estimation |
2024-05-29T01:12:53Z | http://arxiv.org/pdf/2405.18682v1 | http://arxiv.org/abs/2405.18682v1 | Shubham Vatsal, Ayush Singh | 2024-05-29T01:12:53Z | cs.CL, cs.AI, cs.LG | Large language models (LLMs) have shown remarkable performance on many tasks
in different domains. However, their performance in closed-book biomedical
machine reading comprehension (MRC) has not been evaluated in depth. In this
work, we evaluate GPT on four closed-book biomedical MRC benchmarks. We
experiment with different conventional prompting techniques as well as
introduce our own novel prompting method. To solve some of the retrieval
problems inherent to LLMs, we propose a prompting strategy named Implicit
Retrieval Augmented Generation (RAG) that alleviates the need for using vector
databases to retrieve important chunks in traditional RAG setups. Moreover, we
report qualitative assessments on the natural language generation outputs from
our approach. The results show that our new prompting technique is able to get
the best performance in two out of four datasets and ranks second in rest of
them. Experiments show that modern-day LLMs like GPT even in a zero-shot
setting can outperform supervised models, leading to new state-of-the-art
(SoTA) results on two of the benchmarks. | cs.CL | Can GPT Redefine Medical Understanding? Evaluating GPT on Biomedical Machine Reading Comprehension |
2024-05-29T00:36:56Z | http://arxiv.org/pdf/2405.18672v2 | http://arxiv.org/abs/2405.18672v2 | Renyi Qu, Mark Yatskar | 2024-06-02T23:30:46Z | cs.CV, cs.CL | (Renyi Qu's Master's Thesis) Recent advancements in interpretable models for
vision-language tasks have achieved competitive performance; however, their
interpretability often suffers due to the reliance on unstructured text outputs
from large language models (LLMs). This introduces randomness and compromises
both transparency and reliability, which are essential for addressing safety
issues in AI systems. We introduce \texttt{Hi-CoDe} (Hierarchical Concept
Decomposition), a novel framework designed to enhance model interpretability
through structured concept analysis. Our approach consists of two main
components: (1) We use GPT-4 to decompose an input image into a structured
hierarchy of visual concepts, thereby forming a visual concept tree. (2) We
then employ an ensemble of simple linear classifiers that operate on
concept-specific features derived from CLIP to perform classification. Our
approach not only aligns with the performance of state-of-the-art models but
also advances transparency by providing clear insights into the decision-making
process and highlighting the importance of various concepts. This allows for a
detailed analysis of potential failure modes and improves model compactness,
therefore setting a new benchmark in interpretability without compromising the
accuracy. | cs.CV | LLM-based Hierarchical Concept Decomposition for Interpretable Fine-Grained Image Classification |
2024-05-29T00:23:55Z | http://arxiv.org/pdf/2405.18669v2 | http://arxiv.org/abs/2405.18669v2 | Vicky Zayats, Peter Chen, Melissa Ferrari, Dirk Padfield | 2024-05-31T15:42:53Z | cs.LG, cs.AI, cs.CL, eess.AS | Integrating multiple generative foundation models, especially those trained
on different modalities, into something greater than the sum of its parts poses
significant challenges. Two key hurdles are the availability of aligned data
(concepts that contain similar meaning but is expressed differently in
different modalities), and effectively leveraging unimodal representations in
cross-domain generative tasks, without compromising their original unimodal
capabilities.
We propose Zipper, a multi-tower decoder architecture that addresses these
concerns by using cross-attention to flexibly compose multimodal generative
models from independently pre-trained unimodal decoders. In our experiments
fusing speech and text modalities, we show the proposed architecture performs
very competitively in scenarios with limited aligned text-speech data. We also
showcase the flexibility of our model to selectively maintain unimodal (e.g.,
text-to-text generation) generation performance by freezing the corresponding
modal tower (e.g. text). In cross-modal tasks such as automatic speech
recognition (ASR) where the output modality is text, we show that freezing the
text backbone results in negligible performance degradation. In cross-modal
tasks such as text-to-speech generation (TTS) where the output modality is
speech, we show that using a pre-trained speech backbone results in superior
performance to the baseline. | cs.LG | Zipper: A Multi-Tower Decoder Architecture for Fusing Modalities |
2024-05-28T23:54:44Z | http://arxiv.org/pdf/2405.18662v1 | http://arxiv.org/abs/2405.18662v1 | Mina Arzaghi, Florian Carichon, Golnoosh Farnadi | 2024-05-28T23:54:44Z | cs.CL, cs.CY, cs.LG | Large Language Models (LLMs) are increasingly integrated into critical
decision-making processes, such as loan approvals and visa applications, where
inherent biases can lead to discriminatory outcomes. In this paper, we examine
the nuanced relationship between demographic attributes and socioeconomic
biases in LLMs, a crucial yet understudied area of fairness in LLMs. We
introduce a novel dataset of one million English sentences to systematically
quantify socioeconomic biases across various demographic groups. Our findings
reveal pervasive socioeconomic biases in both established models such as GPT-2
and state-of-the-art models like Llama 2 and Falcon. We demonstrate that these
biases are significantly amplified when considering intersectionality, with
LLMs exhibiting a remarkable capacity to extract multiple demographic
attributes from names and then correlate them with specific socioeconomic
biases. This research highlights the urgent necessity for proactive and robust
bias mitigation techniques to safeguard against discriminatory outcomes when
deploying these powerful models in critical real-world applications. | cs.CL | Understanding Intrinsic Socioeconomic Biases in Large Language Models |
2024-05-28T23:32:46Z | http://arxiv.org/pdf/2405.18653v1 | http://arxiv.org/abs/2405.18653v1 | Yutao Yang, Jie Zhou, Xuanwen Ding, Tianyu Huai, Shunyu Liu, Qin Chen, Liang He, Yuan Xie | 2024-05-28T23:32:46Z | cs.CL | Recently, foundation language models (LMs) have marked significant
achievements in the domains of natural language processing (NLP) and computer
vision (CV). Unlike traditional neural network models, foundation LMs obtain a
great ability for transfer learning by acquiring rich commonsense knowledge
through pre-training on extensive unsupervised datasets with a vast number of
parameters. However, they still can not emulate human-like continuous learning
due to catastrophic forgetting. Consequently, various continual learning
(CL)-based methodologies have been developed to refine LMs, enabling them to
adapt to new tasks without forgetting previous knowledge. However, a systematic
taxonomy of existing approaches and a comparison of their performance are still
lacking, which is the gap that our survey aims to fill. We delve into a
comprehensive review, summarization, and classification of the existing
literature on CL-based approaches applied to foundation language models, such
as pre-trained language models (PLMs), large language models (LLMs) and
vision-language models (VLMs). We divide these studies into offline CL and
online CL, which consist of traditional methods, parameter-efficient-based
methods, instruction tuning-based methods and continual pre-training methods.
Offline CL encompasses domain-incremental learning, task-incremental learning,
and class-incremental learning, while online CL is subdivided into hard task
boundary and blurry task boundary settings. Additionally, we outline the
typical datasets and metrics employed in CL research and provide a detailed
analysis of the challenges and future work for LMs-based continual learning. | cs.CL | Recent Advances of Foundation Language Models-based Continual Learning: A Survey |
2024-05-28T23:28:28Z | http://arxiv.org/pdf/2406.02577v1 | http://arxiv.org/abs/2406.02577v1 | Suraj Anand, David Getzen | 2024-05-28T23:28:28Z | cs.CL, cs.CR, cs.LG | Numerous algorithms have been proposed to $\textit{align}$ language models to
remove undesirable behaviors. However, the challenges associated with a very
large state space and creating a proper reward function often result in various
jailbreaks. Our paper aims to examine this effect of reward in the controlled
setting of positive sentiment language generation. Instead of online training
of a reward model based on human feedback, we employ a statically learned
sentiment classifier. We also consider a setting where our model's weights and
activations are exposed to an end-user after training. We examine a pretrained
GPT-2 through the lens of mechanistic interpretability before and after
proximal policy optimization (PPO) has been applied to promote positive
sentiment responses. Using these insights, we (1) attempt to "hack" the PPO-ed
model to generate negative sentiment responses and (2) add a term to the reward
function to try and alter `negative' weights. | cs.CL | Are PPO-ed Language Models Hackable? |
2024-05-28T23:20:24Z | http://arxiv.org/pdf/2405.18649v1 | http://arxiv.org/abs/2405.18649v1 | Nan Jiang, Xiaopeng Li, Shiqi Wang, Qiang Zhou, Soneya Binta Hossain, Baishakhi Ray, Varun Kumar, Xiaofei Ma, Anoop Deoras | 2024-05-28T23:20:24Z | cs.CL, cs.AI, cs.SE | In the domain of code generation, self-debugging is crucial. It allows LLMs
to refine their generated code based on execution feedback. This is
particularly important because generating correct solutions in one attempt
proves challenging for complex tasks. Prior works on self-debugging mostly
focus on prompting methods by providing LLMs with few-shot examples, which work
poorly on small open-sourced LLMs. In this work, we propose a training
framework that significantly improves self-debugging capability of LLMs.
Intuitively, we observe that a chain of explanations on the wrong code followed
by code refinement helps LLMs better analyze the wrong code and do refinement.
We thus propose an automated pipeline to collect a high-quality dataset for
code explanation and refinement by generating a number of explanations and
refinement trajectories and filtering via execution verification. We perform
supervised fine-tuning (SFT) and further reinforcement learning (RL) on both
success and failure trajectories with a novel reward design considering code
explanation and refinement quality. SFT improves the pass@1 by up to 15.92% and
pass@10 by 9.30% over four benchmarks. RL training brings additional up to
3.54% improvement on pass@1 and 2.55% improvement on pass@10. The trained LLMs
show iterative refinement ability, and can keep refining code continuously.
Lastly, our human evaluation shows that the LLMs trained with our framework
generate more useful code explanations and help developers better understand
bugs in source code. | cs.CL | Training LLMs to Better Self-Debug and Explain Code |
2024-05-28T23:01:57Z | http://arxiv.org/pdf/2405.18642v1 | http://arxiv.org/abs/2405.18642v1 | Xiaobo Guo, Jay Desai, Srinivasan H. Sengamedu | 2024-05-28T23:01:57Z | cs.AI, cs.CL | To generate summaries that include multiple aspects or topics for text
documents, most approaches use clustering or topic modeling to group relevant
sentences and then generate a summary for each group. These approaches struggle
to optimize the summarization and clustering algorithms jointly. On the other
hand, aspect-based summarization requires known aspects. Our solution
integrates topic discovery and summarization into a single step. Given text
data, our Joint Aspect Discovery and Summarization algorithm (JADS) discovers
aspects from the input and generates a summary of the topics, in one step. We
propose a self-supervised framework that creates a labeled dataset by first
mixing sentences from multiple documents (e.g., CNN/DailyMail articles) as the
input and then uses the article summaries from the mixture as the labels. The
JADS model outperforms the two-step baselines. With pretraining, the model
achieves better performance and stability. Furthermore, embeddings derived from
JADS exhibit superior clustering capabilities. Our proposed method achieves
higher semantic alignment with ground truth and is factual. | cs.AI | JADS: A Framework for Self-supervised Joint Aspect Discovery and Summarization |
2024-05-28T22:48:53Z | http://arxiv.org/pdf/2405.18639v1 | http://arxiv.org/abs/2405.18639v1 | Brian A. Yuan, Joseph G. Makin | 2024-05-28T22:48:53Z | q-bio.NC, cs.CL, cs.LG, cs.SD, eess.AS | Recent work on intracranial brain-machine interfaces has demonstrated that
spoken speech can be decoded with high accuracy, essentially by treating the
problem as an instance of supervised learning and training deep neural networks
to map from neural activity to text. However, such networks pay for their
expressiveness with very large numbers of labeled data, a requirement that is
particularly burdensome for invasive neural recordings acquired from human
patients. On the other hand, these patients typically produce speech outside of
the experimental blocks used for training decoders. Making use of such data,
and data from other patients, to improve decoding would ease the burden of data
collection -- especially onerous for dys- and anarthric patients. Here we
demonstrate that this is possible, by reengineering wav2vec -- a simple,
self-supervised, fully convolutional model that learns latent representations
of audio using a noise-contrastive loss -- for electrocorticographic (ECoG)
data. We train this model on unlabelled ECoG recordings, and subsequently use
it to transform ECoG from labeled speech sessions into wav2vec's representation
space, before finally training a supervised encoder-decoder to map these
representations to text. We experiment with various numbers of labeled blocks;
for almost all choices, the new representations yield superior decoding
performance to the original ECoG data, and in no cases do they yield worse.
Performance can also be improved in some cases by pretraining wav2vec on
another patient's data. In the best cases, wav2vec's representations decrease
word error rates over the original data by upwards of 50%. | q-bio.NC | Improving Speech Decoding from ECoG with Self-Supervised Pretraining |
2024-05-28T22:45:28Z | http://arxiv.org/pdf/2405.18638v1 | http://arxiv.org/abs/2405.18638v1 | Aparna Elangovan, Ling Liu, Lei Xu, Sravan Bodapati, Dan Roth | 2024-05-28T22:45:28Z | cs.CL, cs.AI | In this position paper, we argue that human evaluation of generative large
language models (LLMs) should be a multidisciplinary undertaking that draws
upon insights from disciplines such as user experience research and human
behavioral psychology to ensure that the experimental design and results are
reliable. The conclusions from these evaluations, thus, must consider factors
such as usability, aesthetics, and cognitive biases. We highlight how cognitive
biases can conflate fluent information and truthfulness, and how cognitive
uncertainty affects the reliability of rating scores such as Likert.
Furthermore, the evaluation should differentiate the capabilities and
weaknesses of increasingly powerful large language models -- which requires
effective test sets. The scalability of human evaluation is also crucial to
wider adoption. Hence, to design an effective human evaluation system in the
age of generative NLP, we propose the ConSiDERS-The-Human evaluation framework
consisting of 6 pillars --Consistency, Scoring Critera, Differentiating, User
Experience, Responsible, and Scalability. | cs.CL | ConSiDERS-The-Human Evaluation Framework: Rethinking Human Evaluation for Generative Large Language Models |
2024-05-28T22:33:02Z | http://arxiv.org/pdf/2405.18634v1 | http://arxiv.org/abs/2405.18634v1 | Yifei Wang, Yuyang Wu, Zeming Wei, Stefanie Jegelka, Yisen Wang | 2024-05-28T22:33:02Z | cs.LG, cs.CL, stat.ML | Going beyond mimicking limited human experiences, recent studies show initial
evidence that, like humans, large language models (LLMs) are capable of
improving their abilities purely by self-correction, i.e., correcting previous
responses through self-examination, in certain circumstances. Nevertheless,
little is known about how such capabilities arise. In this work, based on a
simplified setup akin to an alignment task, we theoretically analyze
self-correction from an in-context learning perspective, showing that when LLMs
give relatively accurate self-examinations as rewards, they are capable of
refining responses in an in-context way. Notably, going beyond previous
theories on over-simplified linear transformers, our theoretical construction
underpins the roles of several key designs of realistic transformers for
self-correction: softmax attention, multi-head attention, and the MLP block. We
validate these findings extensively on synthetic datasets. Inspired by these
findings, we also illustrate novel applications of self-correction, such as
defending against LLM jailbreaks, where a simple self-correction step does make
a large difference. We believe that these findings will inspire further
research on understanding, exploiting, and enhancing self-correction for
building better foundation models. | cs.LG | A Theoretical Understanding of Self-Correction through In-context Alignment |
2024-05-28T22:19:30Z | http://arxiv.org/pdf/2405.18628v2 | http://arxiv.org/abs/2405.18628v2 | Hao Mark Chen, Wayne Luk, Ka Fai Cedric Yiu, Rui Li, Konstantin Mishchenko, Stylianos I. Venieris, Hongxiang Fan | 2024-06-02T14:58:48Z | cs.LG, cs.CL | The auto-regressive decoding of Large Language Models (LLMs) results in
significant overheads in their hardware performance. While recent research has
investigated various speculative decoding techniques for multi-token
generation, these efforts have primarily focused on improving processing speed
such as throughput. Crucially, they often neglect other metrics essential for
real-life deployments, such as memory consumption and training cost. To
overcome these limitations, we propose a novel parallel prompt decoding that
requires only $0.0002$% trainable parameters, enabling efficient training on a
single A100-40GB GPU in just 16 hours. Inspired by the human natural language
generation process, $PPD$ approximates outputs generated at future timesteps in
parallel by using multiple prompt tokens. This approach partially recovers the
missing conditional dependency information necessary for multi-token
generation, resulting in up to a 28% higher acceptance rate for long-range
predictions. Furthermore, we present a hardware-aware dynamic sparse tree
technique that adaptively optimizes this decoding scheme to fully leverage the
computational capacities on different GPUs. Through extensive experiments
across LLMs ranging from MobileLlama to Vicuna-13B on a wide range of
benchmarks, our approach demonstrates up to 2.49$\times$ speedup and maintains
a minimal runtime memory overhead of just $0.0004$%. More importantly, our
parallel prompt decoding can serve as an orthogonal optimization for
synergistic integration with existing speculative decoding, showing up to
$1.22\times$ further speed improvement. Our code is available at
https://github.com/hmarkc/parallel-prompt-decoding. | cs.LG | Hardware-Aware Parallel Prompt Decoding for Memory-Efficient Acceleration of LLM Inference |
2024-05-28T21:59:56Z | http://arxiv.org/pdf/2405.18620v1 | http://arxiv.org/abs/2405.18620v1 | Aditya Gunturu, Shivesh Jadon, Nandi Zhang, Jarin Thundathil, Wesley Willett, Ryo Suzuki | 2024-05-28T21:59:56Z | cs.HC, cs.AI, cs.CL | We introduce RealitySummary, a mixed reality reading assistant that can
enhance any printed or digital document using on-demand text extraction,
summarization, and augmentation. While augmented reading tools promise to
enhance physical reading experiences with overlaid digital content, prior
systems have typically required pre-processed documents, which limits their
generalizability and real-world use cases. In this paper, we explore on-demand
document augmentation by leveraging large language models. To understand
generalizable techniques for diverse documents, we first conducted an
exploratory design study which identified five categories of document
enhancements (summarization, augmentation, navigation, comparison, and
extraction). Based on this, we developed a proof-of-concept system that can
automatically extract and summarize text using Google Cloud OCR and GPT-4, then
embed information around documents using a Microsoft Hololens 2 and Apple
Vision Pro. We demonstrate real-time examples of six specific document
augmentations: 1) summaries, 2) comparison tables, 3) timelines, 4) keyword
lists, 5) summary highlighting, and 6) information cards. Results from a
usability study (N=12) and in-the-wild study (N=11) highlight the potential
benefits of on-demand MR document enhancement and opportunities for future
research. | cs.HC | RealitySummary: On-Demand Mixed Reality Document Enhancement using Large Language Models |
2024-05-28T21:42:35Z | http://arxiv.org/pdf/2405.18613v1 | http://arxiv.org/abs/2405.18613v1 | Ali Hürriyetoğlu, Osman Mutlu, Fırat Duruşan, Erdem Yörük | 2024-05-28T21:42:35Z | cs.CL, cs.CY, cs.DB, cs.LG | GLOCON is a database of contentious events automatically extracted from
national news sources from various countries in multiple languages. National
news sources are utilized, and complete news archives are processed to create
an event list for each source. Automation is achieved using a gold standard
corpus sampled randomly from complete news archives (Y\"or\"uk et al. 2022) and
all annotated by at least two domain experts based on the event definition
provided in Duru\c{s}an et al. (2022). | cs.CL | GLOCON Database: Design Decisions and User Manual (v1.0) |
2024-05-28T21:38:20Z | http://arxiv.org/pdf/2406.00053v2 | http://arxiv.org/abs/2406.00053v2 | Suraj Anand, Michael A. Lepori, Jack Merullo, Ellie Pavlick | 2024-07-01T18:23:43Z | cs.CL, cs.LG | Language models have the ability to perform in-context learning (ICL),
allowing them to flexibly adapt their behavior based on context. This contrasts
with in-weights learning, where information is statically encoded in model
parameters from iterated observations of the data. Despite this apparent
ability to learn in-context, language models are known to struggle when faced
with unseen or rarely seen tokens. Hence, we study $\textbf{structural
in-context learning}$, which we define as the ability of a model to execute
in-context learning on arbitrary tokens -- so called because the model must
generalize on the basis of e.g. sentence structure or task structure, rather
than semantic content encoded in token embeddings. An ideal model would be able
to do both: flexibly deploy in-weights operations (in order to robustly
accommodate ambiguous or unknown contexts using encoded semantic information)
and structural in-context operations (in order to accommodate novel tokens). We
study structural in-context algorithms in a simple part-of-speech setting using
both practical and toy models. We find that active forgetting, a technique that
was recently introduced to help models generalize to new languages, forces
models to adopt structural in-context learning solutions. Finally, we introduce
$\textbf{temporary forgetting}$, a straightforward extension of active
forgetting that enables one to control how much a model relies on in-weights
vs. in-context solutions. Importantly, temporary forgetting allows us to induce
a $\textit{dual process strategy}$ where in-context and in-weights solutions
coexist within a single model. | cs.CL | Dual Process Learning: Controlling Use of In-Context vs. In-Weights Strategies with Weight Forgetting |
2024-05-28T21:34:01Z | http://arxiv.org/pdf/2405.18605v1 | http://arxiv.org/abs/2405.18605v1 | Bridget T. McInnes, Jiawei Tang, Darshini Mahendran, Mai H. Nguyen | 2024-05-28T21:34:01Z | cs.CL, cs.IR, q-bio.MN | This paper presents a methodology for enhancing relation extraction from
biomedical texts, focusing specifically on chemical-gene interactions.
Leveraging the BioBERT model and a multi-layer fully connected network
architecture, our approach integrates the ChemProt and DrugProt datasets using
a novel merging strategy. Through extensive experimentation, we demonstrate
significant performance improvements, particularly in CPR groups shared between
the datasets. The findings underscore the importance of dataset merging in
augmenting sample counts and improving model accuracy. Moreover, the study
highlights the potential of automated information extraction in biomedical
research and clinical practice. | cs.CL | BioBERT-based Deep Learning and Merged ChemProt-DrugProt for Enhanced Biomedical Relation Extraction |
2024-05-28T20:43:53Z | http://arxiv.org/pdf/2405.18572v1 | http://arxiv.org/abs/2405.18572v1 | Saswat Das, Marco Romanelli, Cuong Tran, Zarreen Reza, Bhavya Kailkhura, Ferdinando Fioretto | 2024-05-28T20:43:53Z | cs.LG, cs.AI, cs.CL | Low-rank approximation techniques have become the de facto standard for
fine-tuning Large Language Models (LLMs) due to their reduced computational and
memory requirements. This paper investigates the effectiveness of these methods
in capturing the shift of fine-tuning datasets from the initial pre-trained
data distribution. Our findings reveal that there are cases in which low-rank
fine-tuning falls short in learning such shifts. This, in turn, produces
non-negligible side effects, especially when fine-tuning is adopted for
toxicity mitigation in pre-trained models, or in scenarios where it is
important to provide fair models. Through comprehensive empirical evidence on
several models, datasets, and tasks, we show that low-rank fine-tuning
inadvertently preserves undesirable biases and toxic behaviors. We also show
that this extends to sequential decision-making tasks, emphasizing the need for
careful evaluation to promote responsible LLMs development. | cs.LG | Low-rank finetuning for LLMs: A fairness perspective |
2024-05-28T20:31:59Z | http://arxiv.org/pdf/2407.00039v1 | http://arxiv.org/abs/2407.00039v1 | Diana E. Gherman, Thorsten O. Zander | 2024-05-28T20:31:59Z | q-bio.NC, cs.CL, cs.HC | Moral judgement is a complex human reaction that engages cognitive and
emotional dimensions. While some of the morality neural correlates are known,
it is currently unclear if we can detect moral violation at a single-trial
level. In a pilot study, here we explore the feasibility of moral judgement
decoding from text stimuli with passive brain-computer interfaces. For
effective moral judgement elicitation, we use video-audio affective priming
prior to text stimuli presentation and attribute the text to moral agents. Our
results show that further efforts are necessary to achieve reliable
classification between moral congruency vs. incongruency states. We obtain good
accuracy results for neutral vs. morally-charged trials. With this research, we
try to pave the way towards neuroadaptive human-computer interaction and more
human-compatible large language models (LLMs) | q-bio.NC | Decoding moral judgement from text: a pilot study |
2024-05-28T20:28:07Z | http://arxiv.org/pdf/2405.18570v3 | http://arxiv.org/abs/2405.18570v3 | Abrar Fahim, Alex Murphy, Alona Fyshe | 2024-06-06T20:29:21Z | cs.CV, cs.CL, cs.IR, cs.LG | Multi-modal contrastive models such as CLIP achieve state-of-the-art
performance in zero-shot classification by embedding input images and texts on
a joint representational space. Recently, a modality gap has been reported in
two-encoder contrastive models like CLIP, meaning that the image and text
embeddings reside in disjoint areas of the latent space. Previous studies
suggest that this gap exists due to 1) the cone effect, 2) mismatched pairs in
the dataset, and 3) insufficient training. We show that, even when accounting
for all these factors, and even when using the same modality, the contrastive
loss actually creates a gap during training. As a result, We propose that the
modality gap is inherent to the two-encoder contrastive loss and rename it the
contrastive gap. We present evidence that attributes this contrastive gap to
low uniformity in CLIP space, resulting in embeddings that occupy only a small
portion of the latent space. To close the gap, we adapt the uniformity and
alignment properties of unimodal contrastive loss to the multi-modal setting
and show that simply adding these terms to the CLIP loss distributes the
embeddings more uniformly in the representational space, closing the gap. In
our experiments, we show that the modified representational space achieves
better performance than default CLIP loss in downstream tasks such as zero-shot
image classification and multi-modal arithmetic. | cs.CV | It's Not a Modality Gap: Characterizing and Addressing the Contrastive Gap |
2024-05-28T19:17:48Z | http://arxiv.org/pdf/2405.18542v1 | http://arxiv.org/abs/2405.18542v1 | Francisco de Arriba-Pérez, Silvia García-Méndez, Francisco J. González-Castaño, Enrique Costa-Montenegro | 2024-05-28T19:17:48Z | cs.AI, cs.CL, cs.HC, cs.LG | Previous researchers have proposed intelligent systems for therapeutic
monitoring of cognitive impairments. However, most existing practical
approaches for this purpose are based on manual tests. This raises issues such
as excessive caretaking effort and the white-coat effect. To avoid these
issues, we present an intelligent conversational system for entertaining
elderly people with news of their interest that monitors cognitive impairment
transparently. Automatic chatbot dialogue stages allow assessing content
description skills and detecting cognitive impairment with Machine Learning
algorithms. We create these dialogue flows automatically from updated news
items using Natural Language Generation techniques. The system also infers the
gold standard of the answers to the questions, so it can assess cognitive
capabilities automatically by comparing these answers with the user responses.
It employs a similarity metric with values in [0, 1], in increasing level of
similarity. To evaluate the performance and usability of our approach, we have
conducted field tests with a test group of 30 elderly people in the earliest
stages of dementia, under the supervision of gerontologists. In the
experiments, we have analysed the effect of stress and concentration in these
users. Those without cognitive impairment performed up to five times better. In
particular, the similarity metric varied between 0.03, for stressed and
unfocused participants, and 0.36, for relaxed and focused users. Finally, we
developed a Machine Learning algorithm based on textual analysis features for
automatic cognitive impairment detection, which attained accuracy, F-measure
and recall levels above 80%. We have thus validated the automatic approach to
detect cognitive impairment in elderly people based on entertainment content. | cs.AI | Automatic detection of cognitive impairment in elderly people using an entertainment chatbot with Natural Language Processing capabilities |
2024-05-28T19:16:17Z | http://arxiv.org/pdf/2405.18540v1 | http://arxiv.org/abs/2405.18540v1 | Seanie Lee, Minsu Kim, Lynn Cherif, David Dobre, Juho Lee, Sung Ju Hwang, Kenji Kawaguchi, Gauthier Gidel, Yoshua Bengio, Nikolay Malkin, Moksh Jain | 2024-05-28T19:16:17Z | cs.CL, cs.CR, cs.LG | Red-teaming, or identifying prompts that elicit harmful responses, is a
critical step in ensuring the safe and responsible deployment of large language
models (LLMs). Developing effective protection against many modes of attack
prompts requires discovering diverse attacks. Automated red-teaming typically
uses reinforcement learning to fine-tune an attacker language model to generate
prompts that elicit undesirable responses from a target LLM, as measured, for
example, by an auxiliary toxicity classifier. We show that even with explicit
regularization to favor novelty and diversity, existing approaches suffer from
mode collapse or fail to generate effective attacks. As a flexible and
probabilistically principled alternative, we propose to use GFlowNet
fine-tuning, followed by a secondary smoothing phase, to train the attacker
model to generate diverse and effective attack prompts. We find that the
attacks generated by our method are effective against a wide range of target
LLMs, both with and without safety tuning, and transfer well between target
LLMs. Finally, we demonstrate that models safety-tuned using a dataset of
red-teaming prompts generated by our method are robust to attacks from other
RL-based red-teaming approaches. | cs.CL | Learning diverse attacks on large language models for robust red-teaming and safety tuning |
2024-05-28T18:34:53Z | http://arxiv.org/pdf/2406.00050v2 | http://arxiv.org/abs/2406.00050v2 | Xinyi Liu, Pinxin Liu, Hangfeng He | 2024-06-04T14:51:25Z | cs.CL, cs.AI | In this study, we investigate the capabilities and inherent biases of
advanced large language models (LLMs) such as GPT-3.5 and GPT-4 in the context
of debate evaluation. We discover that LLM's performance exceeds humans and
surpasses the performance of state-of-the-art methods fine-tuned on extensive
datasets in debate evaluation. We additionally explore and analyze biases
present in LLMs, including positional bias, lexical bias, order bias, which may
affect their evaluative judgments. Our findings reveal a consistent bias in
both GPT-3.5 and GPT-4 towards the second candidate response presented,
attributed to prompt design. We also uncover lexical biases in both GPT-3.5 and
GPT-4, especially when label sets carry connotations such as numerical or
sequential, highlighting the critical need for careful label verbalizer
selection in prompt design. Additionally, our analysis indicates a tendency of
both models to favor the debate's concluding side as the winner, suggesting an
end-of-discussion bias. | cs.CL | An Empirical Analysis on Large Language Models in Debate Evaluation |
2024-05-28T18:01:52Z | http://arxiv.org/pdf/2405.18492v2 | http://arxiv.org/abs/2405.18492v2 | Felix B Mueller, Rebekka Görge, Anna K Bernzen, Janna C Pirk, Maximilian Poretschkin | 2024-06-28T16:12:39Z | cs.CL, cs.AI, I.2.7 | Memorization in large language models (LLMs) is a growing concern. LLMs have
been shown to easily reproduce parts of their training data, including
copyrighted work. This is an important problem to solve, as it may violate
existing copyright laws as well as the European AI Act. In this work, we
propose a systematic analysis to quantify the extent of potential copyright
infringements in LLMs using European law as an example. Unlike previous work,
we evaluate instruction-finetuned models in a realistic end-user scenario. Our
analysis builds on a proposed threshold of 160 characters, which we borrow from
the German Copyright Service Provider Act and a fuzzy text matching algorithm
to identify potentially copyright-infringing textual reproductions. The
specificity of countermeasures against copyright infringement is analyzed by
comparing model behavior on copyrighted and public domain data. We investigate
what behaviors models show instead of producing protected text (such as refusal
or hallucination) and provide a first legal assessment of these behaviors. We
find that there are huge differences in copyright compliance, specificity, and
appropriate refusal among popular LLMs. Alpaca, GPT 4, GPT 3.5, and Luminous
perform best in our comparison, with OpenGPT-X, Alpaca, and Luminous producing
a particularly low absolute number of potential copyright violations. Code will
be published soon. | cs.CL | LLMs and Memorization: On Quality and Specificity of Copyright Compliance |
2024-05-28T17:59:53Z | http://arxiv.org/pdf/2405.18433v1 | http://arxiv.org/abs/2405.18433v1 | Łukasz Borchmann | 2024-05-28T17:59:53Z | cs.CL | We perform a missing, reproducible evaluation of all publicly available GPT-4
family models concerning the Document Understanding field, where it is
frequently required to comprehend text spacial arrangement and visual clues in
addition to textual semantics. Benchmark results indicate that though it is
hard to achieve satisfactory results with text-only models, GPT-4 Vision Turbo
performs well when one provides both text recognized by an external OCR engine
and document images on the input. Evaluation is followed by analyses that
suggest possible contamination of textual GPT-4 models and indicate the
significant performance drop for lengthy documents. | cs.CL | Notes on Applicability of GPT-4 to Document Understanding |
2024-05-28T17:57:06Z | http://arxiv.org/pdf/2405.18415v1 | http://arxiv.org/abs/2405.18415v1 | Yuhui Zhang, Alyssa Unell, Xiaohan Wang, Dhruba Ghosh, Yuchang Su, Ludwig Schmidt, Serena Yeung-Levy | 2024-05-28T17:57:06Z | cs.CV, cs.AI, cs.CL, cs.LG | Image classification is one of the most fundamental capabilities of machine
vision intelligence. In this work, we revisit the image classification task
using visually-grounded language models (VLMs) such as GPT-4V and LLaVA. We
find that existing proprietary and public VLMs, despite often using CLIP as a
vision encoder and having many more parameters, significantly underperform CLIP
on standard image classification benchmarks like ImageNet. To understand the
reason, we explore several hypotheses concerning the inference algorithms,
training objectives, and data processing in VLMs. Our analysis reveals that the
primary cause is data-related: critical information for image classification is
encoded in the VLM's latent space but can only be effectively decoded with
enough training data. Specifically, there is a strong correlation between the
frequency of class exposure during VLM training and instruction-tuning and the
VLM's performance in those classes; when trained with sufficient data, VLMs can
match the accuracy of state-of-the-art classification models. Based on these
findings, we enhance a VLM by integrating classification-focused datasets into
its training, and demonstrate that the enhanced classification performance of
the VLM transfers to its general capabilities, resulting in an improvement of
11.8% on the newly collected ImageWikiQA dataset. | cs.CV | Why are Visually-Grounded Language Models Bad at Image Classification? |
2024-05-28T17:56:46Z | http://arxiv.org/pdf/2405.18414v1 | http://arxiv.org/abs/2405.18414v1 | Jialin Dong, Bahare Fatemi, Bryan Perozzi, Lin F. Yang, Anton Tsitsulin | 2024-05-28T17:56:46Z | cs.CL, cs.AI, cs.LG, cs.SI | Retrieval Augmented Generation (RAG) has greatly improved the performance of
Large Language Model (LLM) responses by grounding generation with context from
existing documents. These systems work well when documents are clearly relevant
to a question context. But what about when a document has partial information,
or less obvious connections to the context? And how should we reason about
connections between documents? In this work, we seek to answer these two core
questions about RAG generation. We introduce G-RAG, a reranker based on graph
neural networks (GNNs) between the retriever and reader in RAG. Our method
combines both connections between documents and semantic information (via
Abstract Meaning Representation graphs) to provide a context-informed ranker
for RAG. G-RAG outperforms state-of-the-art approaches while having smaller
computational footprint. Additionally, we assess the performance of PaLM 2 as a
reranker and find it to significantly underperform G-RAG. This result
emphasizes the importance of reranking for RAG even when using Large Language
Models. | cs.CL | Don't Forget to Connect! Improving RAG with Graph-based Reranking |
2024-05-28T17:46:36Z | http://arxiv.org/pdf/2405.18406v1 | http://arxiv.org/abs/2405.18406v1 | Jaehong Yoon, Shoubin Yu, Mohit Bansal | 2024-05-28T17:46:36Z | cs.CV, cs.AI, cs.CL | Recent video generative models primarily rely on carefully written text
prompts for specific tasks, like inpainting or style editing. They require
labor-intensive textual descriptions for input videos, hindering their
flexibility to adapt personal/raw videos to user specifications. This paper
proposes RACCooN, a versatile and user-friendly video-to-paragraph-to-video
generative framework that supports multiple video editing capabilities such as
removal, addition, and modification, through a unified pipeline. RACCooN
consists of two principal stages: Video-to-Paragraph (V2P) and
Paragraph-to-Video (P2V). In the V2P stage, we automatically describe video
scenes in well-structured natural language, capturing both the holistic context
and focused object details. Subsequently, in the P2V stage, users can
optionally refine these descriptions to guide the video diffusion model,
enabling various modifications to the input video, such as removing, changing
subjects, and/or adding new objects. The proposed approach stands out from
other methods through several significant contributions: (1) RACCooN suggests a
multi-granular spatiotemporal pooling strategy to generate well-structured
video descriptions, capturing both the broad context and object details without
requiring complex human annotations, simplifying precise video content editing
based on text for users. (2) Our video generative model incorporates
auto-generated narratives or instructions to enhance the quality and accuracy
of the generated content. It supports the addition of video objects,
inpainting, and attribute modification within a unified framework, surpassing
existing video editing and inpainting benchmarks. The proposed framework
demonstrates impressive versatile capabilities in video-to-paragraph
generation, video content editing, and can be incorporated into other SoTA
video generative models for further enhancement. | cs.CV | RACCooN: Remove, Add, and Change Video Content with Auto-Generated Narratives |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.