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
|
|---|---|---|---|---|---|---|---|---|
2024-02-07T13:22:17Z | http://arxiv.org/pdf/2402.04824v1 | http://arxiv.org/abs/2402.04824v1 | Philipp Sadler, Sherzod Hakimov, David Schlangen | 2024-02-07T13:22:17Z | cs.CL | Albrecht and Stone (2018) state that modeling of changing behaviors remains
an open problem "due to the essentially unconstrained nature of what other
agents may do". In this work we evaluate the adaptability of neural artificial
agents towards assumed partner behaviors in a collaborative reference game. In
this game success is achieved when a knowledgeable Guide can verbally lead a
Follower to the selection of a specific puzzle piece among several distractors.
We frame this language grounding and coordination task as a reinforcement
learning problem and measure to which extent a common reinforcement training
algorithm (PPO) is able to produce neural agents (the Guides) that perform well
with various heuristic Follower behaviors that vary along the dimensions of
confidence and autonomy. We experiment with a learning signal that in addition
to the goal condition also respects an assumed communicative effort. Our
results indicate that this novel ingredient leads to communicative strategies
that are less verbose (staying silent in some of the steps) and that with
respect to that the Guide's strategies indeed adapt to the partner's level of
confidence and autonomy. | cs.CL | Learning Communication Policies for Different Follower Behaviors in a Collaborative Reference Game |
2024-02-07T13:01:43Z | http://arxiv.org/pdf/2402.04812v1 | http://arxiv.org/abs/2402.04812v1 | Lois Rink, Job Meijdam, David Graus | 2024-02-07T13:01:43Z | cs.CL | Understanding preferences, opinions, and sentiment of the workforce is
paramount for effective employee lifecycle management. Open-ended survey
responses serve as a valuable source of information. This paper proposes a
machine learning approach for aspect-based sentiment analysis (ABSA) of Dutch
open-ended responses in employee satisfaction surveys. Our approach aims to
overcome the inherent noise and variability in these responses, enabling a
comprehensive analysis of sentiments that can support employee lifecycle
management. Through response clustering we identify six key aspects (salary,
schedule, contact, communication, personal attention, agreements), which we
validate by domain experts. We compile a dataset of 1,458 Dutch survey
responses, revealing label imbalance in aspects and sentiments. We propose
few-shot approaches for ABSA based on Dutch BERT models, and compare them
against bag-of-words and zero-shot baselines. Our work significantly
contributes to the field of ABSA by demonstrating the first successful
application of Dutch pre-trained language models to aspect-based sentiment
analysis in the domain of human resources (HR). | cs.CL | Aspect-Based Sentiment Analysis for Open-Ended HR Survey Responses |
2024-02-07T12:31:13Z | http://arxiv.org/pdf/2402.04792v2 | http://arxiv.org/abs/2402.04792v2 | Shangmin Guo, Biao Zhang, Tianlin Liu, Tianqi Liu, Misha Khalman, Felipe Llinares, Alexandre Rame, Thomas Mesnard, Yao Zhao, Bilal Piot, Johan Ferret, Mathieu Blondel | 2024-02-29T20:59:17Z | cs.AI, cs.CL, cs.HC | Direct alignment from preferences (DAP) methods, such as DPO, have recently
emerged as efficient alternatives to reinforcement learning from human feedback
(RLHF), that do not require a separate reward model. However, the preference
datasets used in DAP methods are usually collected ahead of training and never
updated, thus the feedback is purely offline. Moreover, responses in these
datasets are often sampled from a language model distinct from the one being
aligned, and since the model evolves over training, the alignment phase is
inevitably off-policy. In this study, we posit that online feedback is key and
improves DAP methods. Our method, online AI feedback (OAIF), uses an LLM as
annotator: on each training iteration, we sample two responses from the current
model and prompt the LLM annotator to choose which one is preferred, thus
providing online feedback. Despite its simplicity, we demonstrate via human
evaluation in several tasks that OAIF outperforms both offline DAP and RLHF
methods. We further show that the feedback leveraged in OAIF is easily
controllable, via instruction prompts to the LLM annotator. | cs.AI | Direct Language Model Alignment from Online AI Feedback |
2024-02-07T12:28:32Z | http://arxiv.org/pdf/2402.04788v3 | http://arxiv.org/abs/2402.04788v3 | Dongping Chen, Ruoxi Chen, Shilin Zhang, Yinuo Liu, Yaochen Wang, Huichi Zhou, Qihui Zhang, Yao Wan, Pan Zhou, Lichao Sun | 2024-06-11T06:21:46Z | cs.CL, cs.AI, cs.CV | Multimodal Large Language Models (MLLMs) have gained significant attention
recently, showing remarkable potential in artificial general intelligence.
However, assessing the utility of MLLMs presents considerable challenges,
primarily due to the absence of multimodal benchmarks that align with human
preferences. Drawing inspiration from the concept of LLM-as-a-Judge within
LLMs, this paper introduces a novel benchmark, termed MLLM-as-a-Judge, to
assess the ability of MLLMs in assisting judges across diverse modalities,
encompassing three distinct tasks: Scoring Evaluation, Pair Comparison, and
Batch Ranking. Our study reveals that, while MLLMs demonstrate remarkable
human-like discernment in Pair Comparison, there is a significant divergence
from human preferences in Scoring Evaluation and Batch Ranking. Furthermore, a
closer examination reveals persistent challenges in the judgment capacities of
LLMs, including diverse biases, hallucinatory responses, and inconsistencies in
judgment, even in advanced models such as GPT-4V. These findings emphasize the
pressing need for enhancements and further research efforts to be undertaken
before regarding MLLMs as fully reliable evaluators. In light of this, we
advocate for additional efforts dedicated to supporting the continuous
development within the domain of MLLM functioning as judges. The code and
dataset are publicly available at our project homepage:
\url{https://mllm-judge.github.io/}. | cs.CL | MLLM-as-a-Judge: Assessing Multimodal LLM-as-a-Judge with Vision-Language Benchmark |
2024-02-07T12:26:12Z | http://arxiv.org/pdf/2402.04787v1 | http://arxiv.org/abs/2402.04787v1 | Marc Braun, Jenny Kunz | 2024-02-07T12:26:12Z | cs.CL | The self-rationalising capabilities of LLMs are appealing because the
generated explanations can give insights into the plausibility of the
predictions. However, how faithful the explanations are to the predictions is
questionable, raising the need to explore the patterns behind them further. To
this end, we propose a hypothesis-driven statistical framework. We use a
Bayesian network to implement a hypothesis about how a task (in our example,
natural language inference) is solved, and its internal states are translated
into natural language with templates. Those explanations are then compared to
LLM-generated free-text explanations using automatic and human evaluations.
This allows us to judge how similar the LLM's and the Bayesian network's
decision processes are. We demonstrate the usage of our framework with an
example hypothesis and two realisations in Bayesian networks. The resulting
models do not exhibit a strong similarity to GPT-3.5. We discuss the
implications of this as well as the framework's potential to approximate LLM
decisions better in future work. | cs.CL | A Hypothesis-Driven Framework for the Analysis of Self-Rationalising Models |
2024-02-07T12:01:02Z | http://arxiv.org/pdf/2402.04779v1 | http://arxiv.org/abs/2402.04779v1 | Qingyu Yin, Xuzheng He, Xiang Zhuang, Yu Zhao, Jianhua Yao, Xiaoyu Shen, Qiang Zhang | 2024-02-07T12:01:02Z | cs.CL, cs.AI | The decoder-only Transformer architecture with causal masking and relative
position encoding (RPE) has become the de facto choice in language modeling.
Despite its exceptional performance across various tasks, we have identified
two limitations: First, it requires all attention scores to be non-zero and sum
up to 1, even if the current embedding has sufficient self-contained
information. This compels the model to assign disproportional excessive
attention to specific tokens. Second, RPE-based Transformers are not universal
approximators due to their limited capacity at encoding absolute positional
information, which limits their application in position-critical tasks. In this
work, we propose StableMask: a parameter-free method to address both
limitations by refining the causal mask. It introduces pseudo-attention values
to balance attention distributions and encodes absolute positional information
via a progressively decreasing mask ratio. StableMask's effectiveness is
validated both theoretically and empirically, showing significant enhancements
in language models with parameter sizes ranging from 71M to 1.4B across diverse
datasets and encoding methods. We further show that it naturally supports (1)
efficient extrapolation without special tricks such as StreamingLLM and (2)
easy integration with existing attention optimization techniques. | cs.CL | StableMask: Refining Causal Masking in Decoder-only Transformer |
2024-02-07T09:36:54Z | http://arxiv.org/pdf/2402.05147v3 | http://arxiv.org/abs/2402.05147v3 | Baohao Liao, Christian Herold, Shahram Khadivi, Christof Monz | 2024-06-21T14:03:48Z | cs.LG, cs.CL | Memory-efficient finetuning of large language models (LLMs) has recently
attracted huge attention with the increasing size of LLMs, primarily due to the
constraints posed by GPU memory limitations and the effectiveness of these
methods compared to full finetuning. Despite the advancements, current
strategies for memory-efficient finetuning, such as QLoRA, exhibit inconsistent
performance across diverse bit-width quantizations and multifaceted tasks. This
inconsistency largely stems from the detrimental impact of the quantization
process on preserved knowledge, leading to catastrophic forgetting and
undermining the utilization of pretrained models for finetuning purposes. In
this work, we introduce a novel quantization framework, ApiQ, designed to
restore the lost information from quantization by concurrently initializing the
LoRA components and quantizing the weights of LLMs. This approach ensures the
maintenance of the original LLM's activation precision while mitigating the
error propagation from shallower into deeper layers. Through comprehensive
evaluations conducted on a spectrum of language tasks with various LLMs, ApiQ
demonstrably minimizes activation error during quantization. Consequently, it
consistently achieves superior finetuning results across various bit-widths. | cs.LG | ApiQ: Finetuning of 2-Bit Quantized Large Language Model |
2024-02-07T09:09:14Z | http://arxiv.org/pdf/2402.04678v3 | http://arxiv.org/abs/2402.04678v3 | Yu-Neng Chuang, Guanchu Wang, Chia-Yuan Chang, Ruixiang Tang, Shaochen Zhong, Fan Yang, Mengnan Du, Xuanting Cai, Xia Hu | 2024-06-26T07:43:11Z | cs.CL, cs.AI, cs.LG | Large Language Models (LLMs) have become proficient in addressing complex
tasks by leveraging their extensive internal knowledge and reasoning
capabilities. However, the black-box nature of these models complicates the
task of explaining their decision-making processes. While recent advancements
demonstrate the potential of leveraging LLMs to self-explain their predictions
through natural language (NL) explanations, their explanations may not
accurately reflect the LLMs' decision-making process due to a lack of fidelity
optimization on the derived explanations. Measuring the fidelity of NL
explanations is a challenging issue, as it is difficult to manipulate the input
context to mask the semantics of these explanations. To this end, we introduce
FaithLM to explain the decision of LLMs with NL explanations. Specifically,
FaithLM designs a method for evaluating the fidelity of NL explanations by
incorporating the contrary explanations to the query process. Moreover, FaithLM
conducts an iterative process to improve the fidelity of derived explanations.
Experiment results on three datasets from multiple domains demonstrate that
FaithLM can significantly improve the fidelity of derived explanations, which
also provides a better alignment with the ground-truth explanations. | cs.CL | FaithLM: Towards Faithful Explanations for Large Language Models |
2024-02-07T09:09:09Z | http://arxiv.org/pdf/2402.04677v1 | http://arxiv.org/abs/2402.04677v1 | Yoshi Suhara, Dimitris Alikaniotis | 2024-02-07T09:09:09Z | cs.CL | Neural abstractive summarization models make summaries in an end-to-end
manner, and little is known about how the source information is actually
converted into summaries. In this paper, we define input sentences that contain
essential information in the generated summary as $\textit{source sentences}$
and study how abstractive summaries are made by analyzing the source sentences.
To this end, we annotate source sentences for reference summaries and system
summaries generated by PEGASUS on document-summary pairs sampled from the
CNN/DailyMail and XSum datasets. We also formulate automatic source sentence
detection and compare multiple methods to establish a strong baseline for the
task. Experimental results show that the perplexity-based method performs well
in highly abstractive settings, while similarity-based methods perform robustly
in relatively extractive settings. Our code and data are available at
https://github.com/suhara/sourcesum. | cs.CL | Source Identification in Abstractive Summarization |
2024-02-07T07:39:27Z | http://arxiv.org/pdf/2402.04636v1 | http://arxiv.org/abs/2402.04636v1 | Roman Koshkin, Katsuhito Sudoh, Satoshi Nakamura | 2024-02-07T07:39:27Z | cs.CL | Decoder-only large language models (LLMs) have recently demonstrated
impressive capabilities in text generation and reasoning. Nonetheless, they
have limited applications in simultaneous machine translation (SiMT), currently
dominated by encoder-decoder transformers. This study demonstrates that, after
fine-tuning on a small dataset comprising causally aligned source and target
sentence pairs, a pre-trained open-source LLM can control input segmentation
directly by generating a special "wait" token. This obviates the need for a
separate policy and enables the LLM to perform English-German and
English-Russian SiMT tasks with BLEU scores that are comparable to those of
specific state-of-the-art baselines. We also evaluated closed-source models
such as GPT-4, which displayed encouraging results in performing the SiMT task
without prior training (zero-shot), indicating a promising avenue for enhancing
future SiMT systems. | cs.CL | TransLLaMa: LLM-based Simultaneous Translation System |
2024-02-07T07:28:34Z | http://arxiv.org/pdf/2402.04631v1 | http://arxiv.org/abs/2402.04631v1 | Mengqi Chen, Bin Guo, Hao Wang, Haoyu Li, Qian Zhao, Jingqi Liu, Yasan Ding, Yan Pan, Zhiwen Yu | 2024-02-07T07:28:34Z | cs.CL | Persuasion, as one of the crucial abilities in human communication, has
garnered extensive attention from researchers within the field of intelligent
dialogue systems. We humans tend to persuade others to change their viewpoints,
attitudes or behaviors through conversations in various scenarios (e.g.,
persuasion for social good, arguing in online platforms). Developing dialogue
agents that can persuade others to accept certain standpoints is essential to
achieving truly intelligent and anthropomorphic dialogue system. Benefiting
from the substantial progress of Large Language Models (LLMs), dialogue agents
have acquired an exceptional capability in context understanding and response
generation. However, as a typical and complicated cognitive psychological
system, persuasive dialogue agents also require knowledge from the domain of
cognitive psychology to attain a level of human-like persuasion. Consequently,
the cognitive strategy-enhanced persuasive dialogue agent (defined as
CogAgent), which incorporates cognitive strategies to achieve persuasive
targets through conversation, has become a predominant research paradigm. To
depict the research trends of CogAgent, in this paper, we first present several
fundamental cognitive psychology theories and give the formalized definition of
three typical cognitive strategies, including the persuasion strategy, the
topic path planning strategy, and the argument structure prediction strategy.
Then we propose a new system architecture by incorporating the formalized
definition to lay the foundation of CogAgent. Representative works are detailed
and investigated according to the combined cognitive strategy, followed by the
summary of authoritative benchmarks and evaluation metrics. Finally, we
summarize our insights on open issues and future directions of CogAgent for
upcoming researchers. | cs.CL | The Future of Cognitive Strategy-enhanced Persuasive Dialogue Agents: New Perspectives and Trends |
2024-02-07T07:24:01Z | http://arxiv.org/pdf/2402.04627v1 | http://arxiv.org/abs/2402.04627v1 | Julio C. Rangel, Tarcisio Mendes de Farias, Ana Claudia Sima, Norio Kobayashi | 2024-02-07T07:24:01Z | cs.AI, cs.CL, cs.DB, cs.IR | The recent success of Large Language Models (LLM) in a wide range of Natural
Language Processing applications opens the path towards novel Question
Answering Systems over Knowledge Graphs leveraging LLMs. However, one of the
main obstacles preventing their implementation is the scarcity of training data
for the task of translating questions into corresponding SPARQL queries,
particularly in the case of domain-specific KGs. To overcome this challenge, in
this study, we evaluate several strategies for fine-tuning the OpenLlama LLM
for question answering over life science knowledge graphs. In particular, we
propose an end-to-end data augmentation approach for extending a set of
existing queries over a given knowledge graph towards a larger dataset of
semantically enriched question-to-SPARQL query pairs, enabling fine-tuning even
for datasets where these pairs are scarce. In this context, we also investigate
the role of semantic "clues" in the queries, such as meaningful variable names
and inline comments. Finally, we evaluate our approach over the real-world Bgee
gene expression knowledge graph and we show that semantic clues can improve
model performance by up to 33% compared to a baseline with random variable
names and no comments included. | cs.AI | SPARQL Generation: an analysis on fine-tuning OpenLLaMA for Question Answering over a Life Science Knowledge Graph |
2024-02-07T07:14:11Z | http://arxiv.org/pdf/2402.04624v2 | http://arxiv.org/abs/2402.04624v2 | Yu Wang, Yifan Gao, Xiusi Chen, Haoming Jiang, Shiyang Li, Jingfeng Yang, Qingyu Yin, Zheng Li, Xian Li, Bing Yin, Jingbo Shang, Julian McAuley | 2024-05-26T23:06:32Z | cs.CL | Existing Large Language Models (LLMs) usually remain static after deployment,
which might make it hard to inject new knowledge into the model. We aim to
build models containing a considerable portion of self-updatable parameters,
enabling the model to integrate new knowledge effectively and efficiently. To
this end, we introduce MEMORYLLM, a model that comprises a transformer and a
fixed-size memory pool within the latent space of the transformer. MEMORYLLM
can self-update with text knowledge and memorize the knowledge injected
earlier. Our evaluations demonstrate the ability of MEMORYLLM to effectively
incorporate new knowledge, as evidenced by its performance on model editing
benchmarks. Meanwhile, the model exhibits long-term information retention
capacity, which is validated through our custom-designed evaluations and
long-context benchmarks. MEMORYLLM also shows operational integrity without any
sign of performance degradation even after nearly a million memory updates. Our
code and model are open-sourced at https://github.com/wangyu-ustc/MemoryLLM. | cs.CL | MEMORYLLM: Towards Self-Updatable Large Language Models |
2024-02-07T06:50:42Z | http://arxiv.org/pdf/2402.04617v2 | http://arxiv.org/abs/2402.04617v2 | Chaojun Xiao, Pengle Zhang, Xu Han, Guangxuan Xiao, Yankai Lin, Zhengyan Zhang, Zhiyuan Liu, Maosong Sun | 2024-05-28T12:05:12Z | cs.CL, cs.AI, cs.LG | Large language models (LLMs) have emerged as a cornerstone in real-world
applications with lengthy streaming inputs (e.g., LLM-driven agents). However,
existing LLMs, pre-trained on sequences with a restricted maximum length,
cannot process longer sequences due to the out-of-domain and distraction
issues. Common solutions often involve continual pre-training on longer
sequences, which will introduce expensive computational overhead and
uncontrollable change in model capabilities. In this paper, we unveil the
intrinsic capacity of LLMs for understanding extremely long sequences without
any fine-tuning. To this end, we introduce a training-free memory-based method,
InfLLM. Specifically, InfLLM stores distant contexts into additional memory
units and employs an efficient mechanism to lookup token-relevant units for
attention computation. Thereby, InfLLM allows LLMs to efficiently process long
sequences with a limited context window and well capture long-distance
dependencies. Without any training, InfLLM enables LLMs that are pre-trained on
sequences consisting of a few thousand tokens to achieve comparable performance
with competitive baselines that continually train these LLMs on long sequences.
Even when the sequence length is scaled to $1,024$K, InfLLM still effectively
captures long-distance dependencies. Our code can be found in
\url{https://github.com/thunlp/InfLLM}. | cs.CL | InfLLM: Training-Free Long-Context Extrapolation for LLMs with an Efficient Context Memory |
2024-02-07T06:48:24Z | http://arxiv.org/pdf/2402.04616v2 | http://arxiv.org/abs/2402.04616v2 | Yijun Tian, Yikun Han, Xiusi Chen, Wei Wang, Nitesh V. Chawla | 2024-04-01T01:28:48Z | cs.CL, cs.AI, cs.LG | Transferring the reasoning capability from stronger large language models
(LLMs) to smaller ones has been quite appealing, as smaller LLMs are more
flexible to deploy with less expense. Among the existing solutions, knowledge
distillation stands out due to its outstanding efficiency and generalization.
However, existing methods suffer from several drawbacks, including limited
knowledge diversity and the lack of rich contextual information. To solve the
problems and facilitate the learning of compact language models, we propose
TinyLLM, a new knowledge distillation paradigm to learn a small student LLM
from multiple large teacher LLMs. In particular, we encourage the student LLM
to not only generate the correct answers but also understand the rationales
behind these answers. Given that different LLMs possess diverse reasoning
skills, we guide the student model to assimilate knowledge from various teacher
LLMs. We further introduce an in-context example generator and a
teacher-forcing Chain-of-Thought strategy to ensure that the rationales are
accurate and grounded in contextually appropriate scenarios. Extensive
experiments on six datasets across two reasoning tasks demonstrate the
superiority of our method. Results show that TinyLLM can outperform large
teacher LLMs significantly, despite a considerably smaller model size. | cs.CL | TinyLLM: Learning a Small Student from Multiple Large Language Models |
2024-02-07T06:32:50Z | http://arxiv.org/pdf/2402.04614v3 | http://arxiv.org/abs/2402.04614v3 | Chirag Agarwal, Sree Harsha Tanneru, Himabindu Lakkaraju | 2024-03-14T03:48:08Z | cs.CL | Large Language Models (LLMs) are deployed as powerful tools for several
natural language processing (NLP) applications. Recent works show that modern
LLMs can generate self-explanations (SEs), which elicit their intermediate
reasoning steps for explaining their behavior. Self-explanations have seen
widespread adoption owing to their conversational and plausible nature.
However, there is little to no understanding of their faithfulness. In this
work, we discuss the dichotomy between faithfulness and plausibility in SEs
generated by LLMs. We argue that while LLMs are adept at generating plausible
explanations -- seemingly logical and coherent to human users -- these
explanations do not necessarily align with the reasoning processes of the LLMs,
raising concerns about their faithfulness. We highlight that the current trend
towards increasing the plausibility of explanations, primarily driven by the
demand for user-friendly interfaces, may come at the cost of diminishing their
faithfulness. We assert that the faithfulness of explanations is critical in
LLMs employed for high-stakes decision-making. Moreover, we emphasize the need
for a systematic characterization of faithfulness-plausibility requirements of
different real-world applications and ensure explanations meet those needs.
While there are several approaches to improving plausibility, improving
faithfulness is an open challenge. We call upon the community to develop novel
methods to enhance the faithfulness of self explanations thereby enabling
transparent deployment of LLMs in diverse high-stakes settings. | cs.CL | Faithfulness vs. Plausibility: On the (Un)Reliability of Explanations from Large Language Models |
2024-02-07T06:13:14Z | http://arxiv.org/pdf/2402.04609v1 | http://arxiv.org/abs/2402.04609v1 | Zhuang Li, Levon Haroutunian, Raj Tumuluri, Philip Cohen, Gholamreza Haffari | 2024-02-07T06:13:14Z | cs.CL | Post-editing has proven effective in improving the quality of text generated
by large language models (LLMs) such as GPT-3.5 or GPT-4, particularly when
direct updating of their parameters to enhance text quality is infeasible or
expensive. However, relying solely on smaller language models for post-editing
can limit the LLMs' ability to generalize across domains. Moreover, the editing
strategies in these methods are not optimally designed for text-generation
tasks. To address these limitations, we propose a neural programmer-interpreter
approach that preserves the domain generalization ability of LLMs when editing
their output. The editing actions in this framework are specifically devised
for text generation. Extensive experiments demonstrate that the
programmer-interpreter significantly enhances GPT-3.5's performance in logical
form-to-text conversion and low-resource machine translation, surpassing other
state-of-the-art (SOTA) LLM post-editing methods in cross-domain settings. | cs.CL | Improving Cross-Domain Low-Resource Text Generation through LLM Post-Editing: A Programmer-Interpreter Approach |
2024-02-07T05:56:54Z | http://arxiv.org/pdf/2402.04601v2 | http://arxiv.org/abs/2402.04601v2 | Haihui Yang, Xiaojun Quan | 2024-06-02T15:50:40Z | cs.CL, cs.AI | Chinese grammatical error correction (CGEC) faces serious overcorrection
challenges when employing autoregressive generative models such as
sequence-to-sequence (Seq2Seq) models and decoder-only large language models
(LLMs). While previous methods aim to address overcorrection in Seq2Seq models,
they are difficult to adapt to decoder-only LLMs. In this paper, we propose an
alignment-enhanced corrector for the overcorrection problem that applies to
both Seq2Seq models and decoder-only LLMs. Our method first trains a correction
model to generate an initial correction of the source sentence. Then, we
combine the source sentence with the initial correction and feed it through an
alignment model for another round of correction, aiming to enforce the
alignment model to focus on potential overcorrection. Moreover, to enhance the
model's ability to identify nuances, we further explore the reverse alignment
of the source sentence and the initial correction. Finally, we transfer the
alignment knowledge from two alignment models to the correction model,
instructing it on how to avoid overcorrection. Experimental results on three
CGEC datasets demonstrate the effectiveness of our approach in alleviating
overcorrection and improving overall performance. Our code has been made
publicly available. | cs.CL | Alirector: Alignment-Enhanced Chinese Grammatical Error Corrector |
2024-02-07T05:05:53Z | http://arxiv.org/pdf/2402.04588v2 | http://arxiv.org/abs/2402.04588v2 | Haoyu Wang, Shuo Wang, Yukun Yan, Xujia Wang, Zhiyu Yang, Yuzhuang Xu, Zhenghao Liu, Liner Yang, Ning Ding, Xu Han, Zhiyuan Liu, Maosong Sun | 2024-02-18T03:56:45Z | cs.CL | Open-source large language models (LLMs) have gained significant strength
across diverse fields. Nevertheless, the majority of studies primarily
concentrate on English, with only limited exploration into the realm of
multilingual abilities. In this work, we therefore construct an open-source
multilingual supervised fine-tuning dataset. Different from previous works that
simply translate English instructions, we consider both the language-specific
and language-agnostic abilities of LLMs. Firstly, we introduce a
knowledge-grounded data augmentation approach to elicit more language-specific
knowledge of LLMs, improving their ability to serve users from different
countries. Moreover, we find modern LLMs possess strong cross-lingual transfer
capabilities, thus repeatedly learning identical content in various languages
is not necessary. Consequently, we can substantially prune the
language-agnostic supervised fine-tuning (SFT) data without any performance
degradation, making multilingual SFT more efficient. The resulting UltraLink
dataset comprises approximately 1 million samples across five languages (i.e.,
En, Zh, Ru, Fr, Es), and the proposed data construction method can be easily
extended to other languages. UltraLink-LM, which is trained on UltraLink,
outperforms several representative baselines across many tasks. | cs.CL | UltraLink: An Open-Source Knowledge-Enhanced Multilingual Supervised Fine-tuning Dataset |
2024-02-07T03:37:19Z | http://arxiv.org/pdf/2402.04559v2 | http://arxiv.org/abs/2402.04559v2 | Chengxing Xie, Canyu Chen, Feiran Jia, Ziyu Ye, Kai Shu, Adel Bibi, Ziniu Hu, Philip Torr, Bernard Ghanem, Guohao Li | 2024-03-10T13:48:43Z | cs.AI, cs.CL, cs.HC | Large Language Model (LLM) agents have been increasingly adopted as
simulation tools to model humans in applications such as social science.
However, one fundamental question remains: can LLM agents really simulate human
behaviors? In this paper, we focus on one of the most critical behaviors in
human interactions, trust, and aim to investigate whether or not LLM agents can
simulate human trust behaviors. We first find that LLM agents generally exhibit
trust behaviors, referred to as agent trust, under the framework of Trust
Games, which are widely recognized in behavioral economics. Then, we discover
that LLM agents can have high behavioral alignment with humans regarding trust
behaviors, particularly for GPT-4, indicating the feasibility to simulate human
trust behaviors with LLM agents. In addition, we probe into the biases in agent
trust and the differences in agent trust towards agents and humans. We also
explore the intrinsic properties of agent trust under conditions including
advanced reasoning strategies and external manipulations. We further offer
important implications of our discoveries for various scenarios where trust is
paramount. Our study provides new insights into the behaviors of LLM agents and
the fundamental analogy between LLMs and humans. | cs.AI | Can Large Language Model Agents Simulate Human Trust Behaviors? |
2024-02-07T02:59:18Z | http://arxiv.org/pdf/2402.04542v1 | http://arxiv.org/abs/2402.04542v1 | Niraj Pahari, Kazutaka Shimada | 2024-02-07T02:59:18Z | cs.CL | Code-switching entails mixing multiple languages. It is an increasingly
occurring phenomenon in social media texts. Usually, code-mixed texts are
written in a single script, even though the languages involved have different
scripts. Pre-trained multilingual models primarily utilize the data in the
native script of the language. In existing studies, the code-switched texts are
utilized as they are. However, using the native script for each language can
generate better representations of the text owing to the pre-trained knowledge.
Therefore, a cross-language-script knowledge sharing architecture utilizing the
cross attention and alignment of the representations of text in individual
language scripts was proposed in this study. Experimental results on two
different datasets containing Nepali-English and Hindi-English code-switched
texts, demonstrate the effectiveness of the proposed method. The interpretation
of the model using model explainability technique illustrates the sharing of
language-specific knowledge between language-specific representations. | cs.CL | Share What You Already Know: Cross-Language-Script Transfer and Alignment for Sentiment Detection in Code-Mixed Data |
2024-02-07T02:06:48Z | http://arxiv.org/pdf/2402.04523v1 | http://arxiv.org/abs/2402.04523v1 | Ryutaro Asahara, Masaki Takahashi, Chiho Iwahashi, Michimasa Inaba | 2024-02-07T02:06:48Z | cs.LG, cs.CL | Chat dialogues contain considerable useful information about a speaker's
interests, preferences, and experiences.Thus, knowledge from open-domain chat
dialogue can be used to personalize various systems and offer recommendations
for advanced information.This study proposed a novel framework SumRec for
recommending information from open-domain chat dialogue.The study also examined
the framework using ChatRec, a newly constructed dataset for training and
evaluation. To extract the speaker and item characteristics, the SumRec
framework employs a large language model (LLM) to generate a summary of the
speaker information from a dialogue and to recommend information about an item
according to the type of user.The speaker and item information are then input
into a score estimation model, generating a recommendation score.Experimental
results show that the SumRec framework provides better recommendations than the
baseline method of using dialogues and item descriptions in their original
form. Our dataset and code is publicly available at
https://github.com/Ryutaro-A/SumRec | cs.LG | SumRec: A Framework for Recommendation using Open-Domain Dialogue |
2024-02-07T01:46:50Z | http://arxiv.org/pdf/2402.04513v3 | http://arxiv.org/abs/2402.04513v3 | Lunyiu Nie, Zhimin Ding, Erdong Hu, Christopher Jermaine, Swarat Chaudhuri | 2024-06-17T18:54:36Z | cs.LG, cs.CL | Large Language Models (LLMs) have a natural role in answering complex queries
about data streams, but the high computational cost of LLM inference makes them
infeasible in many such tasks. We propose online cascade learning, the first
approach to address this challenge. The objective here is to learn a "cascade"
of models, starting with lower-capacity models (such as logistic regression)
and ending with a powerful LLM, along with a deferral policy that determines
the model to be used on a given input. We formulate the task of learning
cascades online as an imitation-learning problem, where smaller models are
updated over time imitating the collected LLM demonstrations, and give a
no-regret algorithm for the problem. Experimental results across four
benchmarks show that our method parallels LLMs in accuracy while cutting down
inference costs by as much as 90% with strong robustness against input
distribution shifts, underscoring its efficacy and adaptability in stream
processing. | cs.LG | Online Cascade Learning for Efficient Inference over Streams |
2024-02-07T01:18:55Z | http://arxiv.org/pdf/2402.04505v1 | http://arxiv.org/abs/2402.04505v1 | Kin Ian Lo, Mehrnoosh Sadrzadeh, Shane Mansfield | 2024-02-07T01:18:55Z | cs.CL, quant-ph | Sheaves are mathematical objects consisting of a base which constitutes a
topological space and the data associated with each open set thereof, e.g.
continuous functions defined on the open sets. Sheaves have originally been
used in algebraic topology and logic. Recently, they have also modelled events
such as physical experiments and natural language disambiguation processes. We
extend the latter models from lexical ambiguities to discourse ambiguities
arising from anaphora. To begin, we calculated a new measure of contextuality
for a dataset of basic anaphoric discourses, resulting in a higher proportion
of contextual models--82.9%--compared to previous work which only yielded 3.17%
contextual models. Then, we show how an extension of the natural language
processing challenge, known as the Winograd Schema, which involves anaphoric
ambiguities can be modelled on the Bell-CHSH scenario with a contextual
fraction of 0.096. | cs.CL | Developments in Sheaf-Theoretic Models of Natural Language Ambiguities |
2024-02-07T00:45:31Z | http://arxiv.org/pdf/2402.04497v1 | http://arxiv.org/abs/2402.04497v1 | Josh Alman, Zhao Song | 2024-02-07T00:45:31Z | cs.LG, cs.CC, cs.CL, cs.DS | Large language models (LLMs) have made fundamental contributions over the
last a few years. To train an LLM, one needs to alternatingly run `forward'
computations and `backward' computations. The forward computation can be viewed
as attention function evaluation, and the backward computation can be viewed as
a gradient computation. In previous work by [Alman and Song, NeurIPS 2023], it
was proved that the forward step can be performed in almost-linear time in
certain parameter regimes, but that there is no truly sub-quadratic time
algorithm in the remaining parameter regimes unless the popular hypothesis SETH
is false. In this work, we show nearly identical results for the harder-seeming
problem of computing the gradient of loss function of one layer attention
network, and thus for the entire process of LLM training. This completely
characterizes the fine-grained complexity of every step of LLM training. | cs.LG | The Fine-Grained Complexity of Gradient Computation for Training Large Language Models |
2024-02-07T00:31:49Z | http://arxiv.org/pdf/2402.04492v1 | http://arxiv.org/abs/2402.04492v1 | Jirayu Burapacheep, Ishan Gaur, Agam Bhatia, Tristan Thrush | 2024-02-07T00:31:49Z | cs.CV, cs.CL | This paper introduces the ColorSwap dataset, designed to assess and improve
the proficiency of multimodal models in matching objects with their colors. The
dataset is comprised of 2,000 unique image-caption pairs, grouped into 1,000
examples. Each example includes a caption-image pair, along with a
``color-swapped'' pair. We follow the Winoground schema: the two captions in an
example have the same words, but the color words have been rearranged to modify
different objects. The dataset was created through a novel blend of automated
caption and image generation with humans in the loop. We evaluate image-text
matching (ITM) and visual language models (VLMs) and find that even the latest
ones are still not robust at this task. GPT-4V and LLaVA score 72% and 42% on
our main VLM metric, although they may improve with more advanced prompting
techniques. On the main ITM metric, contrastive models such as CLIP and SigLIP
perform close to chance (at 12% and 30%, respectively), although the
non-contrastive BLIP ITM model is stronger (87%). We also find that finetuning
on fewer than 2,000 examples yields significant performance gains on this
out-of-distribution word-order understanding task. The dataset is here:
https://github.com/Top34051/colorswap. | cs.CV | ColorSwap: A Color and Word Order Dataset for Multimodal Evaluation |
2024-02-07T00:21:46Z | http://arxiv.org/pdf/2403.09676v1 | http://arxiv.org/abs/2403.09676v1 | Linge Guo | 2024-02-07T00:21:46Z | cs.CL, cs.AI | This research critically navigates the intricate landscape of AI deception,
concentrating on deceptive behaviours of Large Language Models (LLMs). My
objective is to elucidate this issue, examine the discourse surrounding it, and
subsequently delve into its categorization and ramifications. The essay
initiates with an evaluation of the AI Safety Summit 2023 (ASS) and
introduction of LLMs, emphasising multidimensional biases that underlie their
deceptive behaviours.The literature review covers four types of deception
categorised: Strategic deception, Imitation, Sycophancy, and Unfaithful
Reasoning, along with the social implications and risks they entail. Lastly, I
take an evaluative stance on various aspects related to navigating the
persistent challenges of the deceptive AI. This encompasses considerations of
international collaborative governance, the reconfigured engagement of
individuals with AI, proposal of practical adjustments, and specific elements
of digital education. | cs.CL | Unmasking the Shadows of AI: Investigating Deceptive Capabilities in Large Language Models |
2024-02-06T23:52:58Z | http://arxiv.org/pdf/2402.04477v1 | http://arxiv.org/abs/2402.04477v1 | Sil Hamilton | 2024-02-06T23:52:58Z | cs.CL, cs.AI | No two authors write alike. Personal flourishes invoked in written
narratives, from lexicon to rhetorical devices, imply a particular author--what
literary theorists label the implied or virtual author; distinct from the real
author or narrator of a text. Early large language models trained on unfiltered
training sets drawn from a variety of discordant sources yielded incoherent
personalities, problematic for conversational tasks but proving useful for
sampling literature from multiple perspectives. Successes in alignment research
in recent years have allowed researchers to impose subjectively consistent
personae on language models via instruction tuning and reinforcement learning
from human feedback (RLHF), but whether aligned models retain the ability to
model an arbitrary virtual author has received little scrutiny. By studying
4,374 stories sampled from three OpenAI language models, we show successive
versions of GPT-3 suffer from increasing degrees of "mode collapse" whereby
overfitting the model during alignment constrains it from generalizing over
authorship: models suffering from mode collapse become unable to assume a
multiplicity of perspectives. Our method and results are significant for
researchers seeking to employ language models in sociological simulations. | cs.CL | Detecting Mode Collapse in Language Models via Narration |
2024-02-06T23:52:10Z | http://arxiv.org/pdf/2402.04476v2 | http://arxiv.org/abs/2402.04476v2 | Jihyung Kil, Chan Hee Song, Boyuan Zheng, Xiang Deng, Yu Su, Wei-Lun Chao | 2024-03-30T05:18:05Z | cs.CV, cs.AI, cs.CL | Automatic web navigation aims to build a web agent that can follow language
instructions to execute complex and diverse tasks on real-world websites.
Existing work primarily takes HTML documents as input, which define the
contents and action spaces (i.e., actionable elements and operations) of
webpages. Nevertheless, HTML documents may not provide a clear task-related
context for each element, making it hard to select the right (sequence of)
actions. In this paper, we propose to contextualize HTML elements through their
"dual views" in webpage screenshots: each HTML element has its corresponding
bounding box and visual content in the screenshot. We build upon the insight --
web developers tend to arrange task-related elements nearby on webpages to
enhance user experiences -- and propose to contextualize each element with its
neighbor elements, using both textual and visual features. The resulting
representations of HTML elements are more informative for the agent to take
action. We validate our method on the recently released Mind2Web dataset, which
features diverse navigation domains and tasks on real-world websites. Our
method consistently outperforms the baseline in all the scenarios, including
cross-task, cross-website, and cross-domain ones. | cs.CV | Dual-View Visual Contextualization for Web Navigation |
2024-02-06T22:24:56Z | http://arxiv.org/pdf/2402.04442v1 | http://arxiv.org/abs/2402.04442v1 | Olumide Ebenezer Ojo, Olaronke Oluwayemisi Adebanji, Alexander Gelbukh, Hiram Calvo, Anna Feldman | 2024-02-06T22:24:56Z | cs.CL | Effective communication between healthcare providers and patients is crucial
to providing high-quality patient care. In this work, we investigate how
Doctor-written and AI-generated texts in healthcare consultations can be
classified using state-of-the-art embeddings and one-shot classification
systems. By analyzing embeddings such as bag-of-words, character n-grams,
Word2Vec, GloVe, fastText, and GPT2 embeddings, we examine how well our
one-shot classification systems capture semantic information within medical
consultations. Results show that the embeddings are capable of capturing
semantic features from text in a reliable and adaptable manner. Overall,
Word2Vec, GloVe and Character n-grams embeddings performed well, indicating
their suitability for modeling targeted to this task. GPT2 embedding also shows
notable performance, indicating its suitability for models tailored to this
task as well. Our machine learning architectures significantly improved the
quality of health conversations when training data are scarce, improving
communication between patients and healthcare providers. | cs.CL | Evaluating Embeddings for One-Shot Classification of Doctor-AI Consultations |
2024-02-06T22:15:09Z | http://arxiv.org/pdf/2402.04437v4 | http://arxiv.org/abs/2402.04437v4 | Haolun Wu, Ye Yuan, Liana Mikaelyan, Alexander Meulemans, Xue Liu, James Hensman, Bhaskar Mitra | 2024-06-18T22:11:27Z | cs.CL, cs.LG | Recent advances in machine learning have significantly impacted the field of
information extraction, with Language Models (LMs) playing a pivotal role in
extracting structured information from unstructured text. Prior works typically
represent information extraction as triplet-centric and use classical metrics
such as precision and recall for evaluation. We reformulate the task to be
entity-centric, enabling the use of diverse metrics that can provide more
insights from various perspectives. We contribute to the field by introducing
Structured Entity Extraction and proposing the Approximate Entity Set OverlaP
(AESOP) metric, designed to appropriately assess model performance. Later, we
introduce a new model that harnesses the power of LMs for enhanced
effectiveness and efficiency by decomposing the extraction task into multiple
stages. Quantitative and human side-by-side evaluations confirm that our model
outperforms baselines, offering promising directions for future advancements in
structured entity extraction. | cs.CL | Learning to Extract Structured Entities Using Language Models |
2024-02-06T21:14:45Z | http://arxiv.org/pdf/2402.04411v2 | http://arxiv.org/abs/2402.04411v2 | Yiyou Sun, Junjie Hu, Wei Cheng, Haifeng Chen | 2024-06-03T01:40:46Z | cs.CL, cs.LG | This paper introduces the retrieval-augmented large language model with
Definite Finite Automaton (DFA-RAG), a novel framework designed to enhance the
capabilities of conversational agents using large language models (LLMs).
Traditional LLMs face challenges in generating regulated and compliant
responses in special scenarios with predetermined response guidelines, like
emotional support and customer service. Our framework addresses these
challenges by embedding a Definite Finite Automaton (DFA), learned from
training dialogues, within the LLM. This structured approach acts as a semantic
router which enables the LLM to adhere to a deterministic response pathway. The
routing is achieved by the retrieval-augmentation generation (RAG) strategy,
which carefully selects dialogue examples aligned with the current
conversational context. The advantages of DFA-RAG include an interpretable
structure through human-readable DFA, context-aware retrieval for responses in
conversations, and plug-and-play compatibility with existing LLMs. Extensive
benchmarks validate DFA-RAG's effectiveness, indicating its potential as a
valuable contribution to the conversational agent. | cs.CL | DFA-RAG: Conversational Semantic Router for Large Language Model with Definite Finite Automaton |
2024-02-06T21:03:52Z | http://arxiv.org/pdf/2402.04401v1 | http://arxiv.org/abs/2402.04401v1 | Zhaoxuan Tan, Qingkai Zeng, Yijun Tian, Zheyuan Liu, Bing Yin, Meng Jiang | 2024-02-06T21:03:52Z | cs.CL | Personalization in large language models (LLMs) is increasingly important,
aiming to align LLM's interactions, content, and recommendations with
individual user preferences. Recent advances in LLM personalization have
spotlighted effective prompt design, by enriching user queries with
non-parametric knowledge through behavior history retrieval and textual
profiles. However, these approaches were limited due to a lack of model
ownership, resulting in constrained customization and privacy issues. Moreover,
they often failed to accurately capture user behavior patterns, especially in
cases where user data were complex and dynamic. To address these shortcomings,
we introduce One PEFT Per User (OPPU), which employs personalized
parameter-efficient fine-tuning (PEFT) modules, to store user-specific behavior
patterns and preferences. By plugging in users' personal PEFT parameters, they
can own and use their LLMs personally. OPPU integrates parametric user
knowledge in the personal PEFT parameters with the non-parametric knowledge
acquired through retrieval and profile. This integration adapts individual LLMs
to user behavior shifts. Experimental results demonstrate that OPPU
significantly outperforms existing prompt-based methods across seven diverse
tasks in the LaMP benchmark. Further in-depth studies reveal OPPU's enhanced
capabilities in handling user behavior shifts, modeling users at different
active levels, maintaining robustness across various user history formats, and
displaying versatility with different PEFT methods. | cs.CL | Democratizing Large Language Models via Personalized Parameter-Efficient Fine-tuning |
2024-02-06T20:52:12Z | http://arxiv.org/pdf/2402.04396v2 | http://arxiv.org/abs/2402.04396v2 | Albert Tseng, Jerry Chee, Qingyao Sun, Volodymyr Kuleshov, Christopher De Sa | 2024-06-04T04:51:52Z | cs.LG, cs.AI, cs.CL | Post-training quantization (PTQ) reduces the memory footprint of LLMs by
quantizing their weights to low-precision. In this work, we introduce QuIP#, a
weight-only PTQ method that achieves state-of-the-art results in extreme
compression regimes ($\le$ 4 bits per weight) using three novel techniques.
First, QuIP# improves QuIP's (Chee et al., 2023) incoherence processing by
using the randomized Hadamard transform, which is faster and has better
theoretical properties. Second, QuIP# uses vector quantization to take
advantage of the ball-shaped sub-Gaussian distribution that incoherent weights
possess: specifically, we introduce a set of hardware-efficient codebooks based
on the highly symmetric $E_8$ lattice, which achieves the optimal 8-dimension
unit ball packing. Third, QuIP# uses fine-tuning to improve fidelity to the
original model. Our experiments show that QuIP# outperforms existing PTQ
methods, enables new behaviors in PTQ scaling, and supports fast inference. Our
code can be found at https://github.com/Cornell-RelaxML/quip-sharp. | cs.LG | QuIP#: Even Better LLM Quantization with Hadamard Incoherence and Lattice Codebooks |
2024-02-06T20:34:49Z | http://arxiv.org/pdf/2403.05548v1 | http://arxiv.org/abs/2403.05548v1 | Raza Ul Mustafa, Nathalie Japkowicz | 2024-02-06T20:34:49Z | cs.CY, cs.AI, cs.CL, cs.IR, cs.LG, cs.SI | Racism and intolerance on social media contribute to a toxic online
environment which may spill offline to foster hatred, and eventually lead to
physical violence. That is the case with online antisemitism, the specific
category of hatred considered in this study. Tracking antisemitic themes and
their associated terminology over time in online discussions could help monitor
the sentiments of their participants and their evolution, and possibly offer
avenues for intervention that may prevent the escalation of hatred. Due to the
large volume and constant evolution of online traffic, monitoring conversations
manually is impractical. Instead, we propose an automated method that extracts
antisemitic themes and terminology from extremist social media over time and
captures their evolution. Since supervised learning would be too limited for
such a task, we created an unsupervised online machine learning approach that
uses large language models to assess the contextual similarity of posts. The
method clusters similar posts together, dividing, and creating additional
clusters over time when sub-themes emerge from existing ones or new themes
appear. The antisemitic terminology used within each theme is extracted from
the posts in each cluster. Our experiments show that our methodology
outperforms existing baselines and demonstrates the kind of themes and
sub-themes it discovers within antisemitic discourse along with their
associated terminology. We believe that our approach will be useful for
monitoring the evolution of all kinds of hatred beyond antisemitism on social
platforms. | cs.CY | Monitoring the evolution of antisemitic discourse on extremist social media using BERT |
2024-02-06T20:18:32Z | http://arxiv.org/pdf/2402.04373v1 | http://arxiv.org/abs/2402.04373v1 | Alakananda Mitra, Saraju P. Mohanty, Elias Kougianos | 2024-02-06T20:18:32Z | cs.CY, cs.CL | We live in the era of Generative Artificial Intelligence (GenAI). Deepfakes
and Large Language Models (LLMs) are two examples of GenAI. Deepfakes, in
particular, pose an alarming threat to society as they are capable of spreading
misinformation and changing the truth. LLMs are powerful language models that
generate general-purpose language. However due to its generative aspect, it can
also be a risk for people if used with ill intentions. The ethical use of these
technologies is a big concern. This short article tries to find out the
interrelationship between them. | cs.CY | The World of Generative AI: Deepfakes and Large Language Models |
2024-02-06T20:11:54Z | http://arxiv.org/pdf/2402.05140v2 | http://arxiv.org/abs/2402.05140v2 | Junhong Shen, Neil Tenenholtz, James Brian Hall, David Alvarez-Melis, Nicolo Fusi | 2024-05-30T17:37:06Z | cs.LG, cs.AI, cs.CL | Large Language Models (LLMs) have demonstrated remarkable proficiency in
understanding and generating natural language. However, their capabilities wane
in highly specialized domains underrepresented in the pretraining corpus, such
as physical and biomedical sciences. This work explores how to repurpose
general LLMs into effective task solvers for specialized domains. We introduce
a novel, model-agnostic framework for learning custom input tags, which are
parameterized as continuous vectors appended to the LLM's embedding layer, to
condition the LLM. We design two types of input tags: domain tags are used to
delimit specialized representations (e.g., chemical formulas) and provide
domain-relevant context; function tags are used to represent specific functions
(e.g., predicting molecular properties) and compress function-solving
instructions. We develop a three-stage protocol to learn these tags using
auxiliary data and domain knowledge. By explicitly disentangling task domains
from task functions, our method enables zero-shot generalization to unseen
problems through diverse combinations of the input tags. It also boosts LLM's
performance in various specialized domains, such as predicting protein or
chemical properties and modeling drug-target interactions, outperforming expert
models tailored to these tasks. | cs.LG | Tag-LLM: Repurposing General-Purpose LLMs for Specialized Domains |
2024-02-06T19:31:26Z | http://arxiv.org/pdf/2402.04347v1 | http://arxiv.org/abs/2402.04347v1 | Michael Zhang, Kush Bhatia, Hermann Kumbong, Christopher Ré | 2024-02-06T19:31:26Z | cs.LG, cs.CL | Linear attentions have shown potential for improving Transformer efficiency,
reducing attention's quadratic complexity to linear in sequence length. This
holds exciting promise for (1) training linear Transformers from scratch, (2)
"finetuned-conversion" of task-specific Transformers into linear versions that
recover task performance, and (3) "pretrained-conversion" of Transformers such
as large language models into linear versions finetunable on downstream tasks.
However, linear attentions often underperform standard softmax attention in
quality. To close this performance gap, we find prior linear attentions lack
key properties of softmax attention tied to good performance: low-entropy (or
"spiky") weights and dot-product monotonicity. We further observe surprisingly
simple feature maps that retain these properties and match softmax performance,
but are inefficient to compute in linear attention. We thus propose Hedgehog, a
learnable linear attention that retains the spiky and monotonic properties of
softmax attention while maintaining linear complexity. Hedgehog uses simple
trainable MLPs to produce attention weights mimicking softmax attention.
Experiments show Hedgehog recovers over 99% of standard Transformer quality in
train-from-scratch and finetuned-conversion settings, outperforming prior
linear attentions up to 6 perplexity points on WikiText-103 with causal GPTs,
and up to 8.7 GLUE score points on finetuned bidirectional BERTs. Hedgehog also
enables pretrained-conversion. Converting a pretrained GPT-2 into a linear
attention variant achieves state-of-the-art 16.7 perplexity on WikiText-103 for
125M subquadratic decoder models. We finally turn a pretrained Llama-2 7B into
a viable linear attention Llama. With low-rank adaptation, Hedgehog-Llama2 7B
achieves 28.1 higher ROUGE-1 points over the base standard attention model,
where prior linear attentions lead to 16.5 point drops. | cs.LG | The Hedgehog & the Porcupine: Expressive Linear Attentions with Softmax Mimicry |
2024-02-06T19:18:56Z | http://arxiv.org/pdf/2402.04335v1 | http://arxiv.org/abs/2402.04335v1 | Dor Bernsohn, Gil Semo, Yaron Vazana, Gila Hayat, Ben Hagag, Joel Niklaus, Rohit Saha, Kyryl Truskovskyi | 2024-02-06T19:18:56Z | cs.CL, cs.AI, cs.LG | In this study, we focus on two main tasks, the first for detecting legal
violations within unstructured textual data, and the second for associating
these violations with potentially affected individuals. We constructed two
datasets using Large Language Models (LLMs) which were subsequently validated
by domain expert annotators. Both tasks were designed specifically for the
context of class-action cases. The experimental design incorporated fine-tuning
models from the BERT family and open-source LLMs, and conducting few-shot
experiments using closed-source LLMs. Our results, with an F1-score of 62.69\%
(violation identification) and 81.02\% (associating victims), show that our
datasets and setups can be used for both tasks. Finally, we publicly release
the datasets and the code used for the experiments in order to advance further
research in the area of legal natural language processing (NLP). | cs.CL | LegalLens: Leveraging LLMs for Legal Violation Identification in Unstructured Text |
2024-02-06T19:18:04Z | http://arxiv.org/pdf/2402.04333v3 | http://arxiv.org/abs/2402.04333v3 | Mengzhou Xia, Sadhika Malladi, Suchin Gururangan, Sanjeev Arora, Danqi Chen | 2024-06-13T03:42:02Z | cs.CL, cs.AI, cs.LG | Instruction tuning has unlocked powerful capabilities in large language
models (LLMs), effectively using combined datasets to develop generalpurpose
chatbots. However, real-world applications often require a specialized suite of
skills (e.g., reasoning). The challenge lies in identifying the most relevant
data from these extensive datasets to effectively develop specific
capabilities, a setting we frame as targeted instruction tuning. We propose
LESS, an optimizer-aware and practically efficient algorithm to effectively
estimate data influences and perform Low-rank gradiEnt Similarity Search for
instruction data selection. Crucially, LESS adapts existing influence
formulations to work with the Adam optimizer and variable-length instruction
data. LESS first constructs a highly reusable and transferable gradient
datastore with low-dimensional gradient features and then selects examples
based on their similarity to few-shot examples embodying a specific capability.
Experiments show that training on a LESS-selected 5% of the data can often
outperform training on the full dataset across diverse downstream tasks.
Furthermore, the selected data is highly transferable: smaller models can be
leveraged to select useful data for larger models and models from different
families. Our qualitative analysis shows that our method goes beyond surface
form cues to identify data that exemplifies the necessary reasoning skills for
the intended downstream application. | cs.CL | LESS: Selecting Influential Data for Targeted Instruction Tuning |
2024-02-06T19:16:55Z | http://arxiv.org/pdf/2402.05138v1 | http://arxiv.org/abs/2402.05138v1 | Zhenwen Liang, Kehan Guo, Gang Liu, Taicheng Guo, Yujun Zhou, Tianyu Yang, Jiajun Jiao, Renjie Pi, Jipeng Zhang, Xiangliang Zhang | 2024-02-06T19:16:55Z | cs.AI, cs.CL | The paper introduces SceMQA, a novel benchmark for scientific multimodal
question answering at the college entrance level. It addresses a critical
educational phase often overlooked in existing benchmarks, spanning high school
to pre-college levels. SceMQA focuses on core science subjects including
Mathematics, Physics, Chemistry, and Biology. It features a blend of
multiple-choice and free-response formats, ensuring a comprehensive evaluation
of AI models' abilities. Additionally, our benchmark provides specific
knowledge points for each problem and detailed explanations for each answer.
SceMQA also uniquely presents problems with identical contexts but varied
questions to facilitate a more thorough and accurate assessment of reasoning
capabilities. In the experiment, we evaluate both open-source and close-source
state-of-the-art Multimodal Large Language Models (MLLMs), across various
experimental settings. The results show that further research and development
are needed in developing more capable MLLM, as highlighted by only 50% to 60%
accuracy achieved by the strongest models. Our benchmark and analysis will be
available at https://scemqa.github.io/ | cs.AI | SceMQA: A Scientific College Entrance Level Multimodal Question Answering Benchmark |
2024-02-06T19:00:40Z | http://arxiv.org/pdf/2402.04315v2 | http://arxiv.org/abs/2402.04315v2 | Chengyu Huang, Zeqiu Wu, Yushi Hu, Wenya Wang | 2024-05-27T09:32:15Z | cs.CL | While recent Large Language Models (LLMs) have proven useful in answering
user queries, they are prone to hallucination, and their responses often lack
credibility due to missing references to reliable sources. An intuitive
solution to these issues would be to include in-text citations referring to
external documents as evidence. While previous works have directly prompted
LLMs to generate in-text citations, their performances are far from
satisfactory, especially when it comes to smaller LLMs. In this work, we
propose an effective training framework using fine-grained rewards to teach
LLMs to generate highly supportive and relevant citations, while ensuring the
correctness of their responses. We also conduct a systematic analysis of
applying these fine-grained rewards to common LLM training strategies,
demonstrating its advantage over conventional practices. We conduct extensive
experiments on Question Answering (QA) datasets taken from the ALCE benchmark
and validate the model's generalizability using EXPERTQA. On LLaMA-2-7B, the
incorporation of fine-grained rewards achieves the best performance among the
baselines, even surpassing that of GPT-3.5-turbo. | cs.CL | Training Language Models to Generate Text with Citations via Fine-grained Rewards |
2024-02-06T18:59:57Z | http://arxiv.org/pdf/2402.04253v1 | http://arxiv.org/abs/2402.04253v1 | Yu Du, Fangyun Wei, Hongyang Zhang | 2024-02-06T18:59:57Z | cs.CL | We introduce AnyTool, a large language model agent designed to revolutionize
the utilization of a vast array of tools in addressing user queries. We utilize
over 16,000 APIs from Rapid API, operating under the assumption that a subset
of these APIs could potentially resolve the queries. AnyTool primarily
incorporates three elements: an API retriever with a hierarchical structure, a
solver aimed at resolving user queries using a selected set of API candidates,
and a self-reflection mechanism, which re-activates AnyTool if the initial
solution proves impracticable. AnyTool is powered by the function calling
feature of GPT-4, eliminating the need for training external modules. We also
revisit the evaluation protocol introduced by previous works and identify a
limitation in this protocol that leads to an artificially high pass rate. By
revising the evaluation protocol to better reflect practical application
scenarios, we introduce an additional benchmark, termed AnyToolBench.
Experiments across various datasets demonstrate the superiority of our AnyTool
over strong baselines such as ToolLLM and a GPT-4 variant tailored for tool
utilization. For instance, AnyTool outperforms ToolLLM by +35.4% in terms of
average pass rate on ToolBench. Code will be available at
https://github.com/dyabel/AnyTool. | cs.CL | AnyTool: Self-Reflective, Hierarchical Agents for Large-Scale API Calls |
2024-02-06T18:59:30Z | http://arxiv.org/pdf/2402.04251v2 | http://arxiv.org/abs/2402.04251v2 | Jannis Vamvas, Rico Sennrich | 2024-06-03T09:33:14Z | cs.CL | Minimum Bayes Risk (MBR) decoding is a text generation technique that has
been shown to improve the quality of machine translations, but is expensive,
even if a sampling-based approximation is used. Besides requiring a large
number of sampled sequences, it requires the pairwise calculation of a utility
metric, which has quadratic complexity. In this paper, we propose to
approximate pairwise metric scores with scores calculated against aggregated
reference representations. This changes the complexity of utility estimation
from $O(n^2)$ to $O(n)$, while empirically preserving most of the quality gains
of MBR decoding. We release our source code at https://github.com/ZurichNLP/mbr | cs.CL | Linear-time Minimum Bayes Risk Decoding with Reference Aggregation |
2024-02-06T18:59:08Z | http://arxiv.org/pdf/2402.04249v2 | http://arxiv.org/abs/2402.04249v2 | Mantas Mazeika, Long Phan, Xuwang Yin, Andy Zou, Zifan Wang, Norman Mu, Elham Sakhaee, Nathaniel Li, Steven Basart, Bo Li, David Forsyth, Dan Hendrycks | 2024-02-27T04:43:08Z | cs.LG, cs.AI, cs.CL, cs.CV | Automated red teaming holds substantial promise for uncovering and mitigating
the risks associated with the malicious use of large language models (LLMs),
yet the field lacks a standardized evaluation framework to rigorously assess
new methods. To address this issue, we introduce HarmBench, a standardized
evaluation framework for automated red teaming. We identify several desirable
properties previously unaccounted for in red teaming evaluations and
systematically design HarmBench to meet these criteria. Using HarmBench, we
conduct a large-scale comparison of 18 red teaming methods and 33 target LLMs
and defenses, yielding novel insights. We also introduce a highly efficient
adversarial training method that greatly enhances LLM robustness across a wide
range of attacks, demonstrating how HarmBench enables codevelopment of attacks
and defenses. We open source HarmBench at
https://github.com/centerforaisafety/HarmBench. | cs.LG | HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal |
2024-02-06T18:54:07Z | http://arxiv.org/pdf/2402.04247v4 | http://arxiv.org/abs/2402.04247v4 | Xiangru Tang, Qiao Jin, Kunlun Zhu, Tongxin Yuan, Yichi Zhang, Wangchunshu Zhou, Meng Qu, Yilun Zhao, Jian Tang, Zhuosheng Zhang, Arman Cohan, Zhiyong Lu, Mark Gerstein | 2024-06-05T06:13:09Z | cs.CY, cs.AI, cs.CL, cs.LG | Intelligent agents powered by large language models (LLMs) have demonstrated
substantial promise in autonomously conducting experiments and facilitating
scientific discoveries across various disciplines. While their capabilities are
promising, these agents, called scientific LLM agents, also introduce novel
vulnerabilities that demand careful consideration for safety. However, there
exists a notable gap in the literature, as there has been no comprehensive
exploration of these vulnerabilities. This perspective paper fills this gap by
conducting a thorough examination of vulnerabilities in LLM-based agents within
scientific domains, shedding light on potential risks associated with their
misuse and emphasizing the need for safety measures. We begin by providing a
comprehensive overview of the potential risks inherent to scientific LLM
agents, taking into account user intent, the specific scientific domain, and
their potential impact on the external environment. Then, we delve into the
origins of these vulnerabilities and provide a scoping review of the limited
existing works. Based on our analysis, we propose a triadic framework involving
human regulation, agent alignment, and an understanding of environmental
feedback (agent regulation) to mitigate these identified risks. Furthermore, we
highlight the limitations and challenges associated with safeguarding
scientific agents and advocate for the development of improved models, robust
benchmarks, and comprehensive regulations to address these issues effectively. | cs.CY | Prioritizing Safeguarding Over Autonomy: Risks of LLM Agents for Science |
2024-02-06T18:43:48Z | http://arxiv.org/pdf/2402.04236v2 | http://arxiv.org/abs/2402.04236v2 | Ji Qi, Ming Ding, Weihan Wang, Yushi Bai, Qingsong Lv, Wenyi Hong, Bin Xu, Lei Hou, Juanzi Li, Yuxiao Dong, Jie Tang | 2024-05-22T17:10:29Z | cs.CV, cs.CL | Vision-Language Models (VLMs) have demonstrated their broad effectiveness
thanks to extensive training in aligning visual instructions to responses.
However, such training of conclusive alignment leads models to ignore essential
visual reasoning, further resulting in failures in meticulous visual problems
and unfaithful responses. Drawing inspiration from human cognition in solving
visual problems (e.g., marking, zoom in), this paper introduces Chain of
Manipulations, a mechanism that enables VLMs to solve problems step-by-step
with evidence. After training, models can solve various visual problems by
eliciting intrinsic manipulations (e.g., grounding, zoom in) with results
(e.g., boxes, image) actively without involving external tools, while also
allowing users to trace error causes. We study the roadmap to implement this
mechanism, including (1) a flexible design of manipulations upon extensive
analysis, (2) an efficient automated data generation pipeline, (3) a compatible
VLM architecture capable of multi-turn multi-image, and (4) a model training
process for versatile capabilities. With the design, we also manually annotate
6K high-quality samples for the challenging graphical mathematical problems.
Our trained model, \textbf{CogCoM}, equipped with this mechanism with 17B
parameters achieves state-of-the-art performance across 9 benchmarks from 4
categories, demonstrating the effectiveness while preserving the
interpretability. Our code, model weights, and collected data are publicly
available at https://github.com/THUDM/CogCoM. | cs.CV | CogCoM: Train Large Vision-Language Models Diving into Details through Chain of Manipulations |
2024-02-06T18:39:43Z | http://arxiv.org/pdf/2402.04232v2 | http://arxiv.org/abs/2402.04232v2 | Ciaran Regan, Nanami Iwahashi, Shogo Tanaka, Mizuki Oka | 2024-02-07T17:27:09Z | cs.AI, cs.CL | Large Language Models (LLMs) have demonstrated a number of human-like
abilities, however the empathic understanding and emotional state of LLMs is
yet to be aligned to that of humans. In this work, we investigate how the
emotional state of generative LLM agents evolves as they perceive new events,
introducing a novel architecture in which new experiences are compared to past
memories. Through this comparison, the agent gains the ability to understand
new experiences in context, which according to the appraisal theory of emotion
is vital in emotion creation. First, the agent perceives new experiences as
time series text data. After perceiving each new input, the agent generates a
summary of past relevant memories, referred to as the norm, and compares the
new experience to this norm. Through this comparison we can analyse how the
agent reacts to the new experience in context. The PANAS, a test of affect, is
administered to the agent, capturing the emotional state of the agent after the
perception of the new event. Finally, the new experience is then added to the
agents memory to be used in the creation of future norms. By creating multiple
experiences in natural language from emotionally charged situations, we test
the proposed architecture on a wide range of scenarios. The mixed results
suggests that introducing context can occasionally improve the emotional
alignment of the agent, but further study and comparison with human evaluators
is necessary. We hope that this paper is another step towards the alignment of
generative agents. | cs.AI | Can Generative Agents Predict Emotion? |
2024-02-06T18:29:39Z | http://arxiv.org/pdf/2402.04222v3 | http://arxiv.org/abs/2402.04222v3 | Esther Ploeger, Wessel Poelman, Miryam de Lhoneux, Johannes Bjerva | 2024-06-16T07:51:41Z | cs.CL | The NLP research community has devoted increased attention to languages
beyond English, resulting in considerable improvements for multilingual NLP.
However, these improvements only apply to a small subset of the world's
languages. Aiming to extend this, an increasing number of papers aspires to
enhance generalizable multilingual performance across languages. To this end,
linguistic typology is commonly used to motivate language selection, on the
basis that a broad typological sample ought to imply generalization across a
broad range of languages. These selections are often described as being
'typologically diverse'. In this work, we systematically investigate NLP
research that includes claims regarding 'typological diversity'. We find there
are no set definitions or criteria for such claims. We introduce metrics to
approximate the diversity of language selection along several axes and find
that the results vary considerably across papers. Crucially, we show that
skewed language selection can lead to overestimated multilingual performance.
We recommend future work to include an operationalization of 'typological
diversity' that empirically justifies the diversity of language samples. | cs.CL | What is "Typological Diversity" in NLP? |
2024-02-06T17:31:20Z | http://arxiv.org/pdf/2402.04177v1 | http://arxiv.org/abs/2402.04177v1 | Berivan Isik, Natalia Ponomareva, Hussein Hazimeh, Dimitris Paparas, Sergei Vassilvitskii, Sanmi Koyejo | 2024-02-06T17:31:20Z | cs.CL, cs.LG, stat.ML | Scaling laws provide important insights that can guide the design of large
language models (LLMs). Existing work has primarily focused on studying scaling
laws for pretraining (upstream) loss. However, in transfer learning settings,
in which LLMs are pretrained on an unsupervised dataset and then finetuned on a
downstream task, we often also care about the downstream performance. In this
work, we study the scaling behavior in a transfer learning setting, where LLMs
are finetuned for machine translation tasks. Specifically, we investigate how
the choice of the pretraining data and its size affect downstream performance
(translation quality) as judged by two metrics: downstream cross-entropy and
BLEU score. Our experiments indicate that the size of the finetuning dataset
and the distribution alignment between the pretraining and downstream data
significantly influence the scaling behavior. With sufficient alignment, both
downstream cross-entropy and BLEU score improve monotonically with more
pretraining data. In such cases, we show that it is possible to predict the
downstream BLEU score with good accuracy using a log-law. However, there are
also cases where moderate misalignment causes the BLEU score to fluctuate or
get worse with more pretraining, whereas downstream cross-entropy monotonically
improves. By analyzing these observations, we provide new practical insights
for choosing appropriate pretraining data. | cs.CL | Scaling Laws for Downstream Task Performance of Large Language Models |
2024-02-06T17:18:59Z | http://arxiv.org/pdf/2402.04161v1 | http://arxiv.org/abs/2402.04161v1 | Ashok Vardhan Makkuva, Marco Bondaschi, Adway Girish, Alliot Nagle, Martin Jaggi, Hyeji Kim, Michael Gastpar | 2024-02-06T17:18:59Z | cs.LG, cs.CL, cs.IT, math.IT, stat.ML | In recent years, attention-based transformers have achieved tremendous
success across a variety of disciplines including natural languages. A key
ingredient behind their success is the generative pretraining procedure, during
which these models are trained on a large text corpus in an auto-regressive
manner. To shed light on this phenomenon, we propose a new framework that
allows both theory and systematic experiments to study the sequential modeling
capabilities of transformers through the lens of Markov chains. Inspired by the
Markovianity of natural languages, we model the data as a Markovian source and
utilize this framework to systematically study the interplay between the
data-distributional properties, the transformer architecture, the learnt
distribution, and the final model performance. In particular, we theoretically
characterize the loss landscape of single-layer transformers and show the
existence of global minima and bad local minima contingent upon the specific
data characteristics and the transformer architecture. Backed by experiments,
we demonstrate that our theoretical findings are in congruence with the
empirical results. We further investigate these findings in the broader context
of higher order Markov chains and deeper architectures, and outline open
problems in this arena. Code is available at
\url{https://github.com/Bond1995/Markov}. | cs.LG | Attention with Markov: A Framework for Principled Analysis of Transformers via Markov Chains |
2024-02-06T17:18:25Z | http://arxiv.org/pdf/2402.04160v1 | http://arxiv.org/abs/2402.04160v1 | Hao Wang, Lei Sha | 2024-02-06T17:18:25Z | cs.CL | Controllable text generation is a growing field within natural language
generation (NLG) that focuses on producing text that meets specific constraints
in real-world applications. Previous approaches, such as plug-and-play
controllers (PPCs), aimed to steer the properties of generated text in a
flexible manner. However, these methods often compromised the integrity of the
language model's decoding process, resulting in less smooth text generation.
Alternatively, other techniques utilized multiple attribute prompts to align
the generated text with desired attributes, but this approach required prompt
design for each attribute and was dependent on the size of the language model.
This paper introduces a novel method for flexible attribute control in text
generation using pre-trained language models (PLMs). The proposed approach aims
to enhance the fluency of generated text by guiding the generation process with
PPCs. The key idea is to dynamically adjust the distribution of generated text
by modifying prompts, effectively constraining the output space of the language
model and influencing the desired attribute. To enable smooth cooperation
between the PLM and the PPC, our work innovatively proposes a new model
fine-tuning method: Reinforcement Learning with Dynamic Adjust Feedback
(RLDAF).This fine-tuning process adapts a small subset of the language model's
parameters based on the generating actions taken during the PPC control
process. The resulting harmonious collaboration between the PLM and PPC leads
to improved smoothness in text generation during inference. Extensive
experiments were conducted on the SST2 dataset, and the proposed method
outperformed previous approaches in various evaluation metrics, including text
fluency and attribute consistency. | cs.CL | Harnessing the Plug-and-Play Controller by Prompting |
2024-02-06T17:15:33Z | http://arxiv.org/pdf/2402.06665v2 | http://arxiv.org/abs/2402.06665v2 | Tarun Gupta, Wenbo Gong, Chao Ma, Nick Pawlowski, Agrin Hilmkil, Meyer Scetbon, Marc Rigter, Ade Famoti, Ashley Juan Llorens, Jianfeng Gao, Stefan Bauer, Danica Kragic, Bernhard Schölkopf, Cheng Zhang | 2024-04-29T23:21:33Z | cs.AI, cs.CL, cs.LG, cs.RO | Recent advances in foundation models, especially in large multi-modal models
and conversational agents, have ignited interest in the potential of generally
capable embodied agents. Such agents will require the ability to perform new
tasks in many different real-world environments. However, current foundation
models fail to accurately model physical interactions and are therefore
insufficient for Embodied AI. The study of causality lends itself to the
construction of veridical world models, which are crucial for accurately
predicting the outcomes of possible interactions. This paper focuses on the
prospects of building foundation world models for the upcoming generation of
embodied agents and presents a novel viewpoint on the significance of causality
within these. We posit that integrating causal considerations is vital to
facilitating meaningful physical interactions with the world. Finally, we
demystify misconceptions about causality in this context and present our
outlook for future research. | cs.AI | The Essential Role of Causality in Foundation World Models for Embodied AI |
2024-02-06T16:03:57Z | http://arxiv.org/pdf/2402.04110v1 | http://arxiv.org/abs/2402.04110v1 | Erik Weber, Jérôme Rutinowski, Markus Pauly | 2024-02-06T16:03:57Z | cs.CL | ChatGPT is notorious for its intransparent behavior. This paper tries to shed
light on this, providing an in-depth analysis of the dark personality traits
and conspiracy beliefs of GPT-3.5 and GPT-4. Different psychological tests and
questionnaires were employed, including the Dark Factor Test, the Mach-IV
Scale, the Generic Conspiracy Belief Scale, and the Conspiracy Mentality Scale.
The responses were analyzed computing average scores, standard deviations, and
significance tests to investigate differences between GPT-3.5 and GPT-4. For
traits that have shown to be interdependent in human studies, correlations were
considered. Additionally, system roles corresponding to groups that have shown
distinct answering behavior in the corresponding questionnaires were applied to
examine the models' ability to reflect characteristics associated with these
roles in their responses. Dark personality traits and conspiracy beliefs were
not particularly pronounced in either model with little differences between
GPT-3.5 and GPT-4. However, GPT-4 showed a pronounced tendency to believe in
information withholding. This is particularly intriguing given that GPT-4 is
trained on a significantly larger dataset than GPT-3.5. Apparently, in this
case an increased data exposure correlates with a greater belief in the control
of information. An assignment of extreme political affiliations increased the
belief in conspiracy theories. Test sequencing affected the models' responses
and the observed correlations, indicating a form of contextual memory. | cs.CL | Behind the Screen: Investigating ChatGPT's Dark Personality Traits and Conspiracy Beliefs |
2024-02-06T15:59:23Z | http://arxiv.org/pdf/2402.04105v2 | http://arxiv.org/abs/2402.04105v2 | Xuechunzi Bai, Angelina Wang, Ilia Sucholutsky, Thomas L. Griffiths | 2024-05-23T15:50:45Z | cs.CY, cs.CL | Large language models (LLMs) can pass explicit social bias tests but still
harbor implicit biases, similar to humans who endorse egalitarian beliefs yet
exhibit subtle biases. Measuring such implicit biases can be a challenge: as
LLMs become increasingly proprietary, it may not be possible to access their
embeddings and apply existing bias measures; furthermore, implicit biases are
primarily a concern if they affect the actual decisions that these systems
make. We address both challenges by introducing two new measures of bias: LLM
Implicit Bias, a prompt-based method for revealing implicit bias; and LLM
Decision Bias, a strategy to detect subtle discrimination in decision-making
tasks. Both measures are based on psychological research: LLM Implicit Bias
adapts the Implicit Association Test, widely used to study the automatic
associations between concepts held in human minds; and LLM Decision Bias
operationalizes psychological results indicating that relative evaluations
between two candidates, not absolute evaluations assessing each independently,
are more diagnostic of implicit biases. Using these measures, we found
pervasive stereotype biases mirroring those in society in 8 value-aligned
models across 4 social categories (race, gender, religion, health) in 21
stereotypes (such as race and criminality, race and weapons, gender and
science, age and negativity). Our prompt-based LLM Implicit Bias measure
correlates with existing language model embedding-based bias methods, but
better predicts downstream behaviors measured by LLM Decision Bias. These new
prompt-based measures draw from psychology's long history of research into
measuring stereotype biases based on purely observable behavior; they expose
nuanced biases in proprietary value-aligned LLMs that appear unbiased according
to standard benchmarks. | cs.CY | Measuring Implicit Bias in Explicitly Unbiased Large Language Models |
2024-02-06T15:46:31Z | http://arxiv.org/pdf/2402.04088v1 | http://arxiv.org/abs/2402.04088v1 | Bayode Ogunleye, Babitha Dharmaraj | 2024-02-06T15:46:31Z | cs.CL, cs.AI, cs.LG, stat.AP, H.3.3 | The dominance of social media has added to the channels of bullying for
perpetrators. Unfortunately, cyberbullying (CB) is the most prevalent
phenomenon in todays cyber world, and is a severe threat to the mental and
physical health of citizens. This opens the need to develop a robust system to
prevent bullying content from online forums, blogs, and social media platforms
to manage the impact in our society. Several machine learning (ML) algorithms
have been proposed for this purpose. However, their performances are not
consistent due to high class imbalance and generalisation issues. In recent
years, large language models (LLMs) like BERT and RoBERTa have achieved
state-of-the-art (SOTA) results in several natural language processing (NLP)
tasks. Unfortunately, the LLMs have not been applied extensively for CB
detection. In our paper, we explored the use of these models for cyberbullying
(CB) detection. We have prepared a new dataset (D2) from existing studies
(Formspring and Twitter). Our experimental results for dataset D1 and D2 showed
that RoBERTa outperformed other models. | cs.CL | The Use of a Large Language Model for Cyberbullying Detection |
2024-02-06T15:41:49Z | http://arxiv.org/pdf/2402.07932v1 | http://arxiv.org/abs/2402.07932v1 | Nicos Isaak | 2024-02-06T15:41:49Z | cs.HC, cs.AI, cs.CL, I.2.0; I.2.3; I.2.4; I.2.7; I.5.1; I.5.4 | The Winograd Schema Challenge (WSC), a seemingly well-thought-out test for
machine intelligence, has been proposed to shed light on developing systems
that exhibit human behavior. Since its introduction, it aimed to pivot the
focus of the AI community from the technology to the science of AI. While
common and trivial for humans, studies show that it is still challenging for
machines, especially when they have to deal with novel schemas, that is,
well-designed sentences that require the resolving of definite pronouns. As
researchers have become increasingly interested in the challenge itself, this
presumably necessitates the availability of an extensive collection of Winograd
schemas, which goes beyond what human experts can reasonably develop
themselves, especially after proposed ways of utilizing them as novel forms of
CAPTCHAs.
To address this necessity, we propose a novel framework that explicitly
focuses on how humans and machines can collaborate as teammates to design novel
schemas from scratch. This is being accomplished by combining two recent
studies from the literature: i) Winventor, a machine-driven approach for the
development of large amounts of Winograd schemas, albeit not of high quality,
and ii) WinoFlexi, an online crowdsourcing system that allows crowd workers to
develop a limited number of schemas often of similar quality to that of
experts. Our proposal crafts a new road map toward developing a novel
collaborative platform that amplifies human and machine intelligence by
combining their complementary strengths. | cs.HC | A Human-Machine Collaboration Framework for the Development of Schemas |
2024-02-06T15:25:09Z | http://arxiv.org/pdf/2402.04075v1 | http://arxiv.org/abs/2402.04075v1 | Reza Khanmohammadi, Ahmed I Ghanem, Kyle Verdecchia, Ryan Hall, Mohamed Elshaikh, Benjamin Movsas, Hassan Bagher-Ebadian, Indrin Chetty, Mohammad M. Ghassemi, Kundan Thind | 2024-02-06T15:25:09Z | cs.CL | This study introduces a novel teacher-student architecture utilizing Large
Language Models (LLMs) to improve prostate cancer radiotherapy symptom
extraction from clinical notes. Mixtral, the student model, initially extracts
symptoms, followed by GPT-4, the teacher model, which refines prompts based on
Mixtral's performance. This iterative process involved 294 single symptom
clinical notes across 12 symptoms, with up to 16 rounds of refinement per
epoch. Results showed significant improvements in extracting symptoms from both
single and multi-symptom notes. For 59 single symptom notes, accuracy increased
from 0.51 to 0.71, precision from 0.52 to 0.82, recall from 0.52 to 0.72, and
F1 score from 0.49 to 0.73. In 375 multi-symptom notes, accuracy rose from 0.24
to 0.43, precision from 0.6 to 0.76, recall from 0.24 to 0.43, and F1 score
from 0.20 to 0.44. These results demonstrate the effectiveness of advanced
prompt engineering in LLMs for radiation oncology use. | cs.CL | Iterative Prompt Refinement for Radiation Oncology Symptom Extraction Using Teacher-Student Large Language Models |
2024-02-06T15:13:17Z | http://arxiv.org/pdf/2402.04068v3 | http://arxiv.org/abs/2402.04068v3 | Ravi Patel, Angus Brayne, Rogier Hintzen, Daniel Jaroslawicz, Georgiana Neculae, Dane Corneil | 2024-06-18T10:42:54Z | cs.LG, cs.CL | Language models hold incredible promise for enabling scientific discovery by
synthesizing massive research corpora. Many complex scientific research
questions have multiple plausible answers, each supported by evidence of
varying strength. However, existing language models lack the capability to
quantitatively and faithfully compare answer plausibility in terms of
supporting evidence. To address this issue, we introduce Retrieve to Explain
(R2E), a retrieval-based language model. R2E scores and ranks all possible
answers to a research question based on evidence retrieved from a document
corpus. The architecture represents each answer only in terms of its supporting
evidence, with the answer itself masked. This allows us to extend feature
attribution methods, such as Shapley values, to transparently attribute each
answer's score back to its supporting evidence at inference time. The
architecture also allows R2E to incorporate new evidence without retraining,
including non-textual data modalities templated into natural language. We
assess on the challenging task of drug target identification from scientific
literature, a human-in-the-loop process where failures are extremely costly and
explainability is paramount. When predicting whether drug targets will
subsequently be confirmed as efficacious in clinical trials, R2E not only
matches non-explainable literature-based models but also surpasses a
genetics-based target identification approach used throughout the
pharmaceutical industry. | cs.LG | Retrieve to Explain: Evidence-driven Predictions with Language Models |
2024-02-06T14:51:55Z | http://arxiv.org/pdf/2402.04049v1 | http://arxiv.org/abs/2402.04049v1 | Amir Taubenfeld, Yaniv Dover, Roi Reichart, Ariel Goldstein | 2024-02-06T14:51:55Z | cs.CL, cs.AI | Recent advancements in natural language processing, especially the emergence
of Large Language Models (LLMs), have opened exciting possibilities for
constructing computational simulations designed to replicate human behavior
accurately. However, LLMs are complex statistical learners without
straightforward deductive rules, making them prone to unexpected behaviors. In
this study, we highlight the limitations of LLMs in simulating human
interactions, particularly focusing on LLMs' ability to simulate political
debates. Our findings indicate a tendency for LLM agents to conform to the
model's inherent social biases despite being directed to debate from certain
political perspectives. This tendency results in behavioral patterns that seem
to deviate from well-established social dynamics among humans. We reinforce
these observations using an automatic self-fine-tuning method, which enables us
to manipulate the biases within the LLM and demonstrate that agents
subsequently align with the altered biases. These results underscore the need
for further research to develop methods that help agents overcome these biases,
a critical step toward creating more realistic simulations. | cs.CL | Systematic Biases in LLM Simulations of Debates |
2024-02-06T14:24:28Z | http://arxiv.org/pdf/2402.04028v1 | http://arxiv.org/abs/2402.04028v1 | Erion Çano, Dario Lamaj | 2024-02-06T14:24:28Z | cs.CL, cs.AI, cs.LG | The scarcity of available text corpora for low-resource languages like
Albanian is a serious hurdle for research in natural language processing tasks.
This paper introduces AlbNews, a collection of 600 topically labeled news
headlines and 2600 unlabeled ones in Albanian. The data can be freely used for
conducting topic modeling research. We report the initial classification scores
of some traditional machine learning classifiers trained with the AlbNews
samples. These results show that basic models outrun the ensemble learning ones
and can serve as a baseline for future experiments. | cs.CL | AlbNews: A Corpus of Headlines for Topic Modeling in Albanian |
2024-02-06T14:16:32Z | http://arxiv.org/pdf/2402.04023v1 | http://arxiv.org/abs/2402.04023v1 | Jaleh Delfani, Constantin Orasan, Hadeel Saadany, Ozlem Temizoz, Eleanor Taylor-Stilgoe, Diptesh Kanojia, Sabine Braun, Barbara Schouten | 2024-02-06T14:16:32Z | cs.CL | This study explores the use of Google Translate (GT) for translating mental
healthcare (MHealth) information and evaluates its accuracy, comprehensibility,
and implications for multilingual healthcare communication through analysing GT
output in the MHealth domain from English to Persian, Arabic, Turkish,
Romanian, and Spanish. Two datasets comprising MHealth information from the UK
National Health Service website and information leaflets from The Royal College
of Psychiatrists were used. Native speakers of the target languages manually
assessed the GT translations, focusing on medical terminology accuracy,
comprehensibility, and critical syntactic/semantic errors. GT output analysis
revealed challenges in accurately translating medical terminology, particularly
in Arabic, Romanian, and Persian. Fluency issues were prevalent across various
languages, affecting comprehension, mainly in Arabic and Spanish. Critical
errors arose in specific contexts, such as bullet-point formatting,
specifically in Persian, Turkish, and Romanian. Although improvements are seen
in longer-text translations, there remains a need to enhance accuracy in
medical and mental health terminology and fluency, whilst also addressing
formatting issues for a more seamless user experience. The findings highlight
the need to use customised translation engines for Mhealth translation and the
challenges when relying solely on machine-translated medical content,
emphasising the crucial role of human reviewers in multilingual healthcare
communication. | cs.CL | Google Translate Error Analysis for Mental Healthcare Information: Evaluating Accuracy, Comprehensibility, and Implications for Multilingual Healthcare Communication |
2024-02-06T13:26:19Z | http://arxiv.org/pdf/2402.03988v2 | http://arxiv.org/abs/2402.03988v2 | Liang-Hsuan Tseng, En-Pei Hu, Cheng-Han Chiang, Yuan Tseng, Hung-yi Lee, Lin-shan Lee, Shao-Hua Sun | 2024-05-28T17:19:59Z | eess.AS, cs.CL, cs.SD | Unsupervised automatic speech recognition (ASR) aims to learn the mapping
between the speech signal and its corresponding textual transcription without
the supervision of paired speech-text data. A word/phoneme in the speech signal
is represented by a segment of speech signal with variable length and unknown
boundary, and this segmental structure makes learning the mapping between
speech and text challenging, especially without paired data. In this paper, we
propose REBORN,Reinforcement-Learned Boundary Segmentation with Iterative
Training for Unsupervised ASR. REBORN alternates between (1) training a
segmentation model that predicts the boundaries of the segmental structures in
speech signals and (2) training the phoneme prediction model, whose input is
the speech feature segmented by the segmentation model, to predict a phoneme
transcription. Since supervised data for training the segmentation model is not
available, we use reinforcement learning to train the segmentation model to
favor segmentations that yield phoneme sequence predictions with a lower
perplexity. We conduct extensive experiments and find that under the same
setting, REBORN outperforms all prior unsupervised ASR models on LibriSpeech,
TIMIT, and five non-English languages in Multilingual LibriSpeech. We
comprehensively analyze why the boundaries learned by REBORN improve the
unsupervised ASR performance. | eess.AS | REBORN: Reinforcement-Learned Boundary Segmentation with Iterative Training for Unsupervised ASR |
2024-02-06T13:19:53Z | http://arxiv.org/pdf/2402.16874v1 | http://arxiv.org/abs/2402.16874v1 | Julien Pierre Edmond Ghali, Kosuke Shima, Koichi Moriyama, Atsuko Mutoh, Nobuhiro Inuzuka | 2024-02-06T13:19:53Z | cs.IR, cs.AI, cs.CL | In the rapidly changing world of smart technology, searching for documents
has become more challenging due to the rise of advanced language models. These
models sometimes face difficulties, like providing inaccurate information,
commonly known as "hallucination." This research focuses on addressing this
issue through Retrieval-Augmented Generation (RAG), a technique that guides
models to give accurate responses based on real facts. To overcome scalability
issues, the study explores connecting user queries with sophisticated language
models such as BERT and Orca2, using an innovative query optimization process.
The study unfolds in three scenarios: first, without RAG, second, without
additional assistance, and finally, with extra help. Choosing the compact yet
efficient Orca2 7B model demonstrates a smart use of computing resources. The
empirical results indicate a significant improvement in the initial language
model's performance under RAG, particularly when assisted with prompts
augmenters. Consistency in document retrieval across different encodings
highlights the effectiveness of using language model-generated queries. The
introduction of UMAP for BERT further simplifies document retrieval while
maintaining strong results. | cs.IR | Enhancing Retrieval Processes for Language Generation with Augmented Queries |
2024-02-06T13:11:19Z | http://arxiv.org/pdf/2402.05136v1 | http://arxiv.org/abs/2402.05136v1 | Tao Yuan, Xuefei Ning, Dong Zhou, Zhijie Yang, Shiyao Li, Minghui Zhuang, Zheyue Tan, Zhuyu Yao, Dahua Lin, Boxun Li, Guohao Dai, Shengen Yan, Yu Wang | 2024-02-06T13:11:19Z | cs.CL | State-of-the-art large language models (LLMs) are now claiming remarkable
supported context lengths of 256k or even more. In contrast, the average
context lengths of mainstream benchmarks are insufficient (5k-21k), and they
suffer from potential knowledge leakage and inaccurate metrics, resulting in
biased evaluation. This paper introduces LV-Eval, a challenging long-context
benchmark with five length levels (16k, 32k, 64k, 128k, and 256k) reaching up
to 256k words. LV-Eval features two main tasks, single-hop QA and multi-hop QA,
comprising 11 bilingual datasets. The design of LV-Eval has incorporated three
key techniques, namely confusing facts insertion, keyword and phrase
replacement, and keyword-recall-based metric design. The advantages of LV-Eval
include controllable evaluation across different context lengths, challenging
test instances with confusing facts, mitigated knowledge leakage, and more
objective evaluations. We evaluate 10 LLMs on LV-Eval and conduct ablation
studies on the techniques used in LV-Eval construction. The results reveal
that: (i) Commercial LLMs generally outperform open-source LLMs when evaluated
within length levels shorter than their claimed context length. However, their
overall performance is surpassed by open-source LLMs with longer context
lengths. (ii) Extremely long-context LLMs, such as Yi-6B-200k, exhibit a
relatively gentle degradation of performance, but their absolute performances
may not necessarily be higher than those of LLMs with shorter context lengths.
(iii) LLMs' performances can significantly degrade in the presence of confusing
information, especially in the pressure test of "needle in a haystack". (iv)
Issues related to knowledge leakage and inaccurate metrics introduce bias in
evaluation, and these concerns are alleviated in LV-Eval. All datasets and
evaluation codes are released at: https://github.com/infinigence/LVEval. | cs.CL | LV-Eval: A Balanced Long-Context Benchmark with 5 Length Levels Up to 256K |
2024-02-06T12:42:21Z | http://arxiv.org/pdf/2402.03962v3 | http://arxiv.org/abs/2402.03962v3 | Patrick Altmeyer, Andrew M. Demetriou, Antony Bartlett, Cynthia C. S. Liem | 2024-05-31T15:16:21Z | cs.AI, cs.CL | Developments in the field of Artificial Intelligence (AI), and particularly
large language models (LLMs), have created a 'perfect storm' for observing
'sparks' of Artificial General Intelligence (AGI) that are spurious. Like
simpler models, LLMs distill meaningful representations in their latent
embeddings that have been shown to correlate with external variables.
Nonetheless, the correlation of such representations has often been linked to
human-like intelligence in the latter but not the former. We probe models of
varying complexity including random projections, matrix decompositions, deep
autoencoders and transformers: all of them successfully distill information
that can be used to predict latent or external variables and yet none of them
have previously been linked to AGI. We argue and empirically demonstrate that
the finding of meaningful patterns in latent spaces of models cannot be seen as
evidence in favor of AGI. Additionally, we review literature from the social
sciences that shows that humans are prone to seek such patterns and
anthropomorphize. We conclude that both the methodological setup and common
public image of AI are ideal for the misinterpretation that correlations
between model representations and some variables of interest are 'caused' by
the model's understanding of underlying 'ground truth' relationships. We,
therefore, call for the academic community to exercise extra caution, and to be
keenly aware of principles of academic integrity, in interpreting and
communicating about AI research outcomes. | cs.AI | Position: Stop Making Unscientific AGI Performance Claims |
2024-02-06T12:34:15Z | http://arxiv.org/pdf/2402.03957v1 | http://arxiv.org/abs/2402.03957v1 | Shruti Singh, Rishabh Gupta | 2024-02-06T12:34:15Z | cs.CL | Computation of document similarity is a critical task in various NLP domains
that has applications in deduplication, matching, and recommendation.
Traditional approaches for document similarity computation include learning
representations of documents and employing a similarity or a distance function
over the embeddings. However, pairwise similarities and differences are not
efficiently captured by individual representations. Graph representations such
as Joint Concept Interaction Graph (JCIG) represent a pair of documents as a
joint undirected weighted graph. JCIGs facilitate an interpretable
representation of document pairs as a graph. However, JCIGs are undirected, and
don't consider the sequential flow of sentences in documents. We propose two
approaches to model document similarity by representing document pairs as a
directed and sparse JCIG that incorporates sequential information. We propose
two algorithms inspired by Supergenome Sorting and Hamiltonian Path that
replace the undirected edges with directed edges. Our approach also sparsifies
the graph to $O(n)$ edges from JCIG's worst case of $O(n^2)$. We show that our
sparse directed graph model architecture consisting of a Siamese encoder and
GCN achieves comparable results to the baseline on datasets not containing
sequential information and beats the baseline by ten points on an instructional
documents dataset containing sequential information. | cs.CL | Sparse Graph Representations for Procedural Instructional Documents |
2024-02-06T11:54:23Z | http://arxiv.org/pdf/2402.03927v2 | http://arxiv.org/abs/2402.03927v2 | Simone Balloccu, Patrícia Schmidtová, Mateusz Lango, Ondřej Dušek | 2024-02-22T12:32:24Z | cs.CL, cs.AI | Natural Language Processing (NLP) research is increasingly focusing on the
use of Large Language Models (LLMs), with some of the most popular ones being
either fully or partially closed-source. The lack of access to model details,
especially regarding training data, has repeatedly raised concerns about data
contamination among researchers. Several attempts have been made to address
this issue, but they are limited to anecdotal evidence and trial and error.
Additionally, they overlook the problem of \emph{indirect} data leaking, where
models are iteratively improved by using data coming from users. In this work,
we conduct the first systematic analysis of work using OpenAI's GPT-3.5 and
GPT-4, the most prominently used LLMs today, in the context of data
contamination. By analysing 255 papers and considering OpenAI's data usage
policy, we extensively document the amount of data leaked to these models
during the first year after the model's release. We report that these models
have been globally exposed to $\sim$4.7M samples from 263 benchmarks. At the
same time, we document a number of evaluation malpractices emerging in the
reviewed papers, such as unfair or missing baseline comparisons and
reproducibility issues. We release our results as a collaborative project on
https://leak-llm.github.io/, where other researchers can contribute to our
efforts. | cs.CL | Leak, Cheat, Repeat: Data Contamination and Evaluation Malpractices in Closed-Source LLMs |
2024-02-06T11:33:57Z | http://arxiv.org/pdf/2402.03916v2 | http://arxiv.org/abs/2402.03916v2 | Qiang Liu, Xiang Tao, Junfei Wu, Shu Wu, Liang Wang | 2024-02-08T16:09:36Z | cs.IR, cs.CL | In this work, we investigate to use Large Language Models (LLMs) for rumor
detection on social media. However, it is challenging for LLMs to reason over
the entire propagation information on social media, which contains news
contents and numerous comments, due to LLMs may not concentrate on key clues in
the complex propagation information, and have trouble in reasoning when facing
massive and redundant information. Accordingly, we propose an LLM-empowered
Rumor Detection (LeRuD) approach, in which we design prompts to teach LLMs to
reason over important clues in news and comments, and divide the entire
propagation information into a Chain-of-Propagation for reducing LLMs' burden.
We conduct extensive experiments on the Twitter and Weibo datasets, and LeRuD
outperforms several state-of-the-art rumor detection models by 3.2% to 7.7%.
Meanwhile, by applying LLMs, LeRuD requires no data for training, and thus
shows more promising rumor detection ability in few-shot or zero-shot
scenarios. | cs.IR | Can Large Language Models Detect Rumors on Social Media? |
2024-02-06T11:29:44Z | http://arxiv.org/pdf/2402.05135v1 | http://arxiv.org/abs/2402.05135v1 | Zijie Zhong, Yunhui Zhang, Ziyi Chang, Zengchang Qin | 2024-02-06T11:29:44Z | cs.AI, cs.CL, cs.IR, 68T07 | Node Importance Estimation (NIE) is crucial for integrating external
information into Large Language Models through Retriever-Augmented Generation.
Traditional methods, focusing on static, single-graph characteristics, lack
adaptability to new graphs and user-specific requirements. CADReN, our proposed
method, addresses these limitations by introducing a Contextual Anchor (CA)
mechanism. This approach enables the network to assess node importance relative
to the CA, considering both structural and semantic features within Knowledge
Graphs (KGs). Extensive experiments show that CADReN achieves better
performance in cross-graph NIE task, with zero-shot prediction ability. CADReN
is also proven to match the performance of previous models on single-graph NIE
task. Additionally, we introduce and opensource two new datasets, RIC200 and
WK1K, specifically designed for cross-graph NIE research, providing a valuable
resource for future developments in this domain. | cs.AI | CADReN: Contextual Anchor-Driven Relational Network for Controllable Cross-Graphs Node Importance Estimation |
2024-02-06T11:12:09Z | http://arxiv.org/pdf/2402.03900v1 | http://arxiv.org/abs/2402.03900v1 | Dechuan Teng, Chunlin Lu, Xiao Xu, Wanxiang Che, Libo Qin | 2024-02-06T11:12:09Z | cs.CL | Recently, Profile-based Spoken Language Understanding (SLU) has gained
increasing attention, which aims to incorporate various types of supplementary
profile information (i.e., Knowledge Graph, User Profile, Context Awareness) to
eliminate the prevalent ambiguities in user utterances. However, existing
approaches can only separately model different profile information, without
considering their interrelationships or excluding irrelevant and conflicting
information within them. To address the above issues, we introduce a
Heterogeneous Graph Attention Network to perform reasoning across multiple
Profile information, called Pro-HAN. Specifically, we design three types of
edges, denoted as intra-Pro, inter-Pro, and utterance-Pro, to capture
interrelationships among multiple Pros. We establish a new state-of-the-art on
the ProSLU dataset, with an improvement of approximately 8% across all three
metrics. Further analysis experiments also confirm the effectiveness of our
method in modeling multi-source profile information. | cs.CL | Pro-HAN: A Heterogeneous Graph Attention Network for Profile-Based Spoken Language Understanding |
2024-02-06T11:10:35Z | http://arxiv.org/pdf/2402.03898v2 | http://arxiv.org/abs/2402.03898v2 | Jongwoo Ko, Sungnyun Kim, Tianyi Chen, Se-Young Yun | 2024-07-03T04:57:41Z | cs.CL, cs.AI, cs.LG | Knowledge distillation (KD) is widely used for compressing a teacher model to
a smaller student model, reducing its inference cost and memory footprint while
preserving model capabilities. However, current KD methods for auto-regressive
sequence models (e.g., large language models) suffer from missing a
standardized objective function. Moreover, the recent use of student-generated
outputs to address training-inference mismatches has significantly escalated
computational costs. To tackle these issues, we introduce DistiLLM, a more
effective and efficient KD framework for auto-regressive language models.
DistiLLM comprises two components: (1) a novel skew Kullback-Leibler divergence
loss, where we unveil and leverage its theoretical properties, and (2) an
adaptive off-policy approach designed to enhance the efficiency in utilizing
student-generated outputs. Extensive experiments, including
instruction-following tasks, demonstrate the effectiveness of DistiLLM in
building high-performing student models while achieving up to 4.3$\times$
speedup compared to recent KD methods. | cs.CL | DistiLLM: Towards Streamlined Distillation for Large Language Models |
2024-02-06T10:49:28Z | http://arxiv.org/pdf/2402.03887v1 | http://arxiv.org/abs/2402.03887v1 | Carolin Müller-Spitzer, Samira Ochs | 2024-02-06T10:49:28Z | cs.CL | This paper focuses on language change based on shifting social norms, in
particular with regard to the debate on language and gender. It is a recurring
argument in this debate that language develops "naturally" and that "severe
interventions" - such as gender-inclusive language is often claimed to be - in
the allegedly "organic" language system are inappropriate and even "dangerous".
Such interventions are, however, not unprecedented. Socially motivated
processes of language change are neither unusual nor new. We focus in our
contribution on one important political-social space in Germany, the German
Bundestag. Taking other struggles about language and gender in the plenaries of
the Bundestag as a starting point, our article illustrates that language and
gender has been a recurring issue in the German Bundestag since the 1980s. We
demonstrate how this is reflected in linguistic practices of the Bundestag, by
the use of a) designations for gays and lesbians; b) pair forms such as
B\"urgerinnen und B\"urger (female and male citizens); and c) female forms of
addresses and personal nouns ('Pr\"asidentin' in addition to 'Pr\"asident').
Lastly, we will discuss implications of these earlier language battles for the
currently very heated debate about gender-inclusive language, especially
regarding new forms with gender symbols like the asterisk or the colon
(Lehrer*innen, Lehrer:innen; male*female teachers) which are intended to
encompass all gender identities. | cs.CL | Shifting social norms as a driving force for linguistic change: Struggles about language and gender in the German Bundestag |
2024-02-06T10:37:21Z | http://arxiv.org/pdf/2402.03877v2 | http://arxiv.org/abs/2402.03877v2 | Spyridon Mouselinos, Henryk Michalewski, Mateusz Malinowski | 2024-02-14T19:33:19Z | cs.CL, cs.AI | Large Language Models (LLMs) demonstrate ever-increasing abilities in
mathematical and algorithmic tasks, yet their geometric reasoning skills are
underexplored. We investigate LLMs' abilities in constructive geometric
problem-solving one of the most fundamental steps in the development of human
mathematical reasoning. Our work reveals notable challenges that the
state-of-the-art LLMs face in this domain despite many successes in similar
areas. LLMs exhibit biases in target variable selection and struggle with 2D
spatial relationships, often misrepresenting and hallucinating objects and
their placements. To this end, we introduce a framework that formulates an
LLMs-based multi-agents system that enhances their existing reasoning potential
by conducting an internal dialogue. This work underscores LLMs' current
limitations in geometric reasoning and improves geometric reasoning
capabilities through self-correction, collaboration, and diverse role
specializations. | cs.CL | Beyond Lines and Circles: Unveiling the Geometric Reasoning Gap in Large Language Models |
2024-02-06T10:32:34Z | http://arxiv.org/pdf/2402.03870v1 | http://arxiv.org/abs/2402.03870v1 | Carolin Müller-Spitzer, Samira Ochs, Alexander Koplenig, Jan-Oliver Rüdiger, Sascha Wolfer | 2024-02-06T10:32:34Z | cs.CL, stat.AP | Research on gender and language is tightly knitted to social debates on
gender equality and non-discriminatory language use. Psycholinguistic scholars
have made significant contributions in this field. However, corpus-based
studies that investigate these matters within the context of language use are
still rare. In our study, we address the question of how much textual material
would actually have to be changed if non-gender-inclusive texts were rewritten
to be gender-inclusive. This quantitative measure is an important empirical
insight, as a recurring argument against the use of gender-inclusive German is
that it supposedly makes written texts too long and complicated. It is also
argued that gender-inclusive language has negative effects on language
learners. However, such effects are only likely if gender-inclusive texts are
very different from those that are not gender-inclusive. In our
corpus-linguistic study, we manually annotated German press texts to identify
the parts that would have to be changed. Our results show that, on average,
less than 1% of all tokens would be affected by gender-inclusive language. This
small proportion calls into question whether gender-inclusive German presents a
substantial barrier to understanding and learning the language, particularly
when we take into account the potential complexities of interpreting masculine
generics. | cs.CL | Less than one percent of words would be affected by gender-inclusive language in German press texts |
2024-02-06T09:50:08Z | http://arxiv.org/pdf/2402.03848v6 | http://arxiv.org/abs/2402.03848v6 | David Peer, Philemon Schöpf, Volckmar Nebendahl, Alexander Rietzler, Sebastian Stabinger | 2024-06-28T06:49:39Z | cs.CL, cs.AI | Traditionally, discriminative models have been the predominant choice for
tasks like document classification and information extraction. These models
make predictions that fall into a limited number of predefined classes,
facilitating a binary true or false evaluation and enabling the direct
calculation of metrics such as the F1 score. However, recent advancements in
generative large language models (GLLMs) have prompted a shift in the field due
to their enhanced zero-shot capabilities, which eliminate the need for a
downstream dataset and computationally expensive fine-tuning. However,
evaluating GLLMs presents a challenge as the binary true or false evaluation
used for discriminative models is not applicable to the predictions made by
GLLMs.
This paper introduces a new metric for generative models called ANLS* for
evaluating a wide variety of tasks, including information extraction and
classification tasks. The ANLS* metric extends existing ANLS metrics as a
drop-in-replacement and is still compatible with previously reported ANLS
scores. An evaluation of 7 different datasets, and more than 10 different GLLMs
together with 3 different prompting methods using the ANLS* metric is also
provided, demonstrating the importance of the proposed metric.
We also benchmark a novel approach to generate prompts for documents, called
SFT, against other prompting techniques such as LATIN. In 6 out of 7 cases, SFT
outperforms other techniques and improves the state-of-the-art, sometimes by as
much as $10$ percentage points.
Sources are available at https://github.com/deepopinion/anls_star_metric | cs.CL | ANLS* -- A Universal Document Processing Metric for Generative Large Language Models |
2024-02-06T09:26:34Z | http://arxiv.org/pdf/2402.04291v2 | http://arxiv.org/abs/2402.04291v2 | Wei Huang, Yangdong Liu, Haotong Qin, Ying Li, Shiming Zhang, Xianglong Liu, Michele Magno, Xiaojuan Qi | 2024-05-15T13:55:12Z | cs.LG, cs.AI, cs.CL | Pretrained large language models (LLMs) exhibit exceptional general language
processing capabilities but come with significant demands on memory and
computational resources. As a powerful compression technology, binarization can
extremely reduce model weights to a mere 1 bit, lowering the expensive
computation and memory requirements. However, existing quantization techniques
fall short of maintaining LLM performance under ultra-low bit-widths. In
response to this challenge, we present BiLLM, a groundbreaking 1-bit
post-training quantization scheme tailored for pretrained LLMs. Based on the
weight distribution of LLMs, BiLLM first identifies and structurally selects
salient weights, and minimizes the compression loss through an effective binary
residual approximation strategy. Moreover, considering the bell-shaped
distribution of the non-salient weights, we propose an optimal splitting search
to group and binarize them accurately. BiLLM achieving for the first time
high-accuracy inference (e.g. 8.41 perplexity on LLaMA2-70B) with only 1.08-bit
weights across various LLMs families and evaluation metrics, outperforms SOTA
quantization methods of LLM by significant margins. Moreover, BiLLM enables the
binarization process of the LLM with 7 billion weights within 0.5 hours on a
single GPU, demonstrating satisfactory time efficiency. Our code is available
at https://github.com/Aaronhuang-778/BiLLM. | cs.LG | BiLLM: Pushing the Limit of Post-Training Quantization for LLMs |
2024-02-06T09:23:26Z | http://arxiv.org/pdf/2402.03832v1 | http://arxiv.org/abs/2402.03832v1 | Khanh Cao Nguyen, Mike Zhang, Syrielle Montariol, Antoine Bosselut | 2024-02-06T09:23:26Z | cs.CL | Skill Extraction involves identifying skills and qualifications mentioned in
documents such as job postings and resumes. The task is commonly tackled by
training supervised models using a sequence labeling approach with BIO tags.
However, the reliance on manually annotated data limits the generalizability of
such approaches. Moreover, the common BIO setting limits the ability of the
models to capture complex skill patterns and handle ambiguous mentions. In this
paper, we explore the use of in-context learning to overcome these challenges,
on a benchmark of 6 uniformized skill extraction datasets. Our approach
leverages the few-shot learning capabilities of large language models (LLMs) to
identify and extract skills from sentences. We show that LLMs, despite not
being on par with traditional supervised models in terms of performance, can
better handle syntactically complex skill mentions in skill extraction tasks. | cs.CL | Rethinking Skill Extraction in the Job Market Domain using Large Language Models |
2024-02-06T09:10:35Z | http://arxiv.org/pdf/2402.03822v2 | http://arxiv.org/abs/2402.03822v2 | Si Shen, Peijun Shen, Danhao Zhu | 2024-02-24T01:11:14Z | cs.AI, cs.CL, cs.LG | This paper presents RevOrder, a novel technique aimed at improving arithmetic
operations in large language models (LLMs) by reversing the output digits in
addition, subtraction, and n-digit by 1-digit (nD by 1D) multiplication tasks.
Our method significantly reduces the Count of Sequential Intermediate Digits
(CSID) to $\mathcal{O}(1)$, a new metric we introduce to assess equation
complexity. Through comprehensive testing, RevOrder not only achieves perfect
accuracy in basic arithmetic operations but also substantially boosts LLM
performance in division tasks, particularly with large numbers where
traditional models struggle. Implementation of RevOrder is cost-effective for
both training and inference phases. Moreover, applying RevOrder to fine-tune
the LLaMA2-7B model on the GSM8K math task results in a considerable
improvement, reducing equation calculation errors by 46% and increasing overall
scores from 41.6 to 44.4. | cs.AI | RevOrder: A Novel Method for Enhanced Arithmetic in Language Models |
2024-02-06T07:52:30Z | http://arxiv.org/pdf/2402.03782v1 | http://arxiv.org/abs/2402.03782v1 | Fred Philippy, Siwen Guo, Shohreh Haddadan, Cedric Lothritz, Jacques Klein, Tegawendé F. Bissyandé | 2024-02-06T07:52:30Z | cs.CL, cs.AI | Soft Prompt Tuning (SPT) is a parameter-efficient method for adapting
pre-trained language models (PLMs) to specific tasks by inserting learnable
embeddings, or soft prompts, at the input layer of the PLM, without modifying
its parameters. This paper investigates the potential of SPT for cross-lingual
transfer. Unlike previous studies on SPT for cross-lingual transfer that often
fine-tune both the soft prompt and the model parameters, we adhere to the
original intent of SPT by keeping the model parameters frozen and only training
the soft prompt. This does not only reduce the computational cost and storage
overhead of full-model fine-tuning, but we also demonstrate that this very
parameter efficiency intrinsic to SPT can enhance cross-lingual transfer
performance to linguistically distant languages. Moreover, we explore how
different factors related to the prompt, such as the length or its
reparameterization, affect cross-lingual transfer performance. | cs.CL | Soft Prompt Tuning for Cross-Lingual Transfer: When Less is More |
2024-02-06T07:51:54Z | http://arxiv.org/pdf/2402.03780v3 | http://arxiv.org/abs/2402.03780v3 | Géraud Faye, Benjamin Icard, Morgane Casanova, Julien Chanson, François Maine, François Bancilhon, Guillaume Gadek, Guillaume Gravier, Paul Égré | 2024-02-26T14:07:20Z | cs.CL, cs.AI, cs.LG | This paper investigates the language of propaganda and its stylistic
features. It presents the PPN dataset, standing for Propagandist Pseudo-News, a
multisource, multilingual, multimodal dataset composed of news articles
extracted from websites identified as propaganda sources by expert agencies. A
limited sample from this set was randomly mixed with papers from the regular
French press, and their URL masked, to conduct an annotation-experiment by
humans, using 11 distinct labels. The results show that human annotators were
able to reliably discriminate between the two types of press across each of the
labels. We propose different NLP techniques to identify the cues used by the
annotators, and to compare them with machine classification. They include the
analyzer VAGO to measure discourse vagueness and subjectivity, a TF-IDF to
serve as a baseline, and four different classifiers: two RoBERTa-based models,
CATS using syntax, and one XGBoost combining syntactic and semantic features. | cs.CL | Exposing propaganda: an analysis of stylistic cues comparing human annotations and machine classification |
2024-02-06T07:43:07Z | http://arxiv.org/pdf/2402.03776v4 | http://arxiv.org/abs/2402.03776v4 | Shahriar Golchin, Nikhil Garuda, Christopher Impey, Matthew Wenger | 2024-03-01T04:48:41Z | cs.CL, cs.AI | Massive open online courses (MOOCs) unlock the doors to free education for
anyone around the globe with access to a computer and the internet. Despite
this democratization of learning, the massive enrollment in these courses means
it is almost impossible for one instructor to assess every student's writing
assignment. As a result, peer grading, often guided by a straightforward
rubric, is the method of choice. While convenient, peer grading often falls
short in terms of reliability and validity. In this study, using 18 distinct
settings, we explore the feasibility of leveraging large language models (LLMs)
to replace peer grading in MOOCs. Specifically, we focus on two
state-of-the-art LLMs: GPT-4 and GPT-3.5, across three distinct courses:
Introductory Astronomy, Astrobiology, and the History and Philosophy of
Astronomy. To instruct LLMs, we use three different prompts based on a variant
of the zero-shot chain-of-thought (Zero-shot-CoT) prompting technique:
Zero-shot-CoT combined with instructor-provided correct answers; Zero-shot-CoT
in conjunction with both instructor-formulated answers and rubrics; and
Zero-shot-CoT with instructor-offered correct answers and LLM-generated
rubrics. Our results show that Zero-shot-CoT, when integrated with
instructor-provided answers and rubrics, produces grades that are more aligned
with those assigned by instructors compared to peer grading. However, the
History and Philosophy of Astronomy course proves to be more challenging in
terms of grading as opposed to other courses. Finally, our study reveals a
promising direction for automating grading systems for MOOCs, especially in
subjects with well-defined rubrics. | cs.CL | Large Language Models As MOOCs Graders |
2024-02-06T07:40:53Z | http://arxiv.org/pdf/2402.03774v1 | http://arxiv.org/abs/2402.03774v1 | Yufan Zhuang, Liyuan Liu, Chandan Singh, Jingbo Shang, Jianfeng Gao | 2024-02-06T07:40:53Z | cs.LG, cs.AI, cs.CL | Decision trees are renowned for their interpretability capability to achieve
high predictive performance, especially on tabular data. Traditionally, they
are constructed through recursive algorithms, where they partition the data at
every node in a tree. However, identifying the best partition is challenging,
as decision trees optimized for local segments may not bring global
generalization. To address this, we introduce MetaTree, which trains a
transformer-based model on filtered outputs from classical algorithms to
produce strong decision trees for classification. Specifically, we fit both
greedy decision trees and optimized decision trees on a large number of
datasets. We then train MetaTree to produce the trees that achieve strong
generalization performance. This training enables MetaTree to not only emulate
these algorithms, but also to intelligently adapt its strategy according to the
context, thereby achieving superior generalization performance. | cs.LG | Learning a Decision Tree Algorithm with Transformers |
2024-02-06T06:48:46Z | http://arxiv.org/pdf/2402.03757v1 | http://arxiv.org/abs/2402.03757v1 | Tianyang Han, Qing Lian, Rui Pan, Renjie Pi, Jipeng Zhang, Shizhe Diao, Yong Lin, Tong Zhang | 2024-02-06T06:48:46Z | cs.CV, cs.CL, cs.LG | Large language models (LLMs) have recently experienced remarkable progress,
where the advent of multi-modal large language models (MLLMs) has endowed LLMs
with visual capabilities, leading to impressive performances in various
multi-modal tasks. However, those powerful MLLMs such as GPT-4V still fail
spectacularly when presented with certain image and text inputs. In this paper,
we identify a typical class of inputs that baffles MLLMs, which consist of
images that are highly relevant but inconsistent with answers, causing MLLMs to
suffer from hallucination. To quantify the effect, we propose CorrelationQA,
the first benchmark that assesses the hallucination level given spurious
images. This benchmark contains 7,308 text-image pairs across 13 categories.
Based on the proposed CorrelationQA, we conduct a thorough analysis on 9
mainstream MLLMs, illustrating that they universally suffer from this
instinctive bias to varying degrees. We hope that our curated benchmark and
evaluation results aid in better assessments of the MLLMs' robustness in the
presence of misleading images. The resource is available in
https://github.com/MasaiahHan/CorrelationQA. | cs.CV | The Instinctive Bias: Spurious Images lead to Hallucination in MLLMs |
2024-02-06T06:23:12Z | http://arxiv.org/pdf/2402.03744v1 | http://arxiv.org/abs/2402.03744v1 | Chao Chen, Kai Liu, Ze Chen, Yi Gu, Yue Wu, Mingyuan Tao, Zhihang Fu, Jieping Ye | 2024-02-06T06:23:12Z | cs.CL | Knowledge hallucination have raised widespread concerns for the security and
reliability of deployed LLMs. Previous efforts in detecting hallucinations have
been employed at logit-level uncertainty estimation or language-level
self-consistency evaluation, where the semantic information is inevitably lost
during the token-decoding procedure. Thus, we propose to explore the dense
semantic information retained within LLMs' \textbf{IN}ternal \textbf{S}tates
for halluc\textbf{I}nation \textbf{DE}tection (\textbf{INSIDE}). In particular,
a simple yet effective \textbf{EigenScore} metric is proposed to better
evaluate responses' self-consistency, which exploits the eigenvalues of
responses' covariance matrix to measure the semantic consistency/diversity in
the dense embedding space. Furthermore, from the perspective of self-consistent
hallucination detection, a test time feature clipping approach is explored to
truncate extreme activations in the internal states, which reduces
overconfident generations and potentially benefits the detection of
overconfident hallucinations. Extensive experiments and ablation studies are
performed on several popular LLMs and question-answering (QA) benchmarks,
showing the effectiveness of our proposal. | cs.CL | INSIDE: LLMs' Internal States Retain the Power of Hallucination Detection |
2024-02-06T05:58:15Z | http://arxiv.org/pdf/2402.03732v1 | http://arxiv.org/abs/2402.03732v1 | Huiling Tu, Shuo Yu, Vidya Saikrishna, Feng Xia, Karin Verspoor | 2024-02-06T05:58:15Z | cs.AI, cs.CL, cs.DL, cs.LG, 68T09, 68T30, 68P20, I.2.6; I.2.4; H.3.7; H.3.3 | Knowledge graphs (KGs) have garnered significant attention for their vast
potential across diverse domains. However, the issue of outdated facts poses a
challenge to KGs, affecting their overall quality as real-world information
evolves. Existing solutions for outdated fact detection often rely on manual
recognition. In response, this paper presents DEAN (Deep outdatEd fAct
detectioN), a novel deep learning-based framework designed to identify outdated
facts within KGs. DEAN distinguishes itself by capturing implicit structural
information among facts through comprehensive modeling of both entities and
relations. To effectively uncover latent out-of-date information, DEAN employs
a contrastive approach based on a pre-defined Relations-to-Nodes (R2N) graph,
weighted by the number of entities. Experimental results demonstrate the
effectiveness and superiority of DEAN over state-of-the-art baseline methods. | cs.AI | Deep Outdated Fact Detection in Knowledge Graphs |
2024-02-06T05:50:04Z | http://arxiv.org/pdf/2402.03728v1 | http://arxiv.org/abs/2402.03728v1 | Hossein Rajaby Faghihi, Parisa Kordjamshidi | 2024-02-06T05:50:04Z | cs.AI, cs.CL, cs.LG, cs.LO | This paper introduces a novel decision-making framework that promotes
consistency among decisions made by diverse models while utilizing external
knowledge. Leveraging the Integer Linear Programming (ILP) framework, we map
predictions from various models into globally normalized and comparable values
by incorporating information about decisions' prior probability, confidence
(uncertainty), and the models' expected accuracy. Our empirical study
demonstrates the superiority of our approach over conventional baselines on
multiple datasets. | cs.AI | Consistent Joint Decision-Making with Heterogeneous Learning Models |
2024-02-06T05:29:05Z | http://arxiv.org/pdf/2402.03720v1 | http://arxiv.org/abs/2402.03720v1 | Rui Li, Jiwei Li, Jiawei Han, Guoyin Wang | 2024-02-06T05:29:05Z | cs.LG, cs.AI, cs.CL, cs.SI | Text-attributed graphs (TAGs) present unique challenges for direct processing
by Language Learning Models (LLMs), yet their extensive commonsense knowledge
and robust reasoning capabilities offer great promise for node classification
in TAGs. Prior research in this field has grappled with issues such as
over-squashing, heterophily, and ineffective graph information integration,
further compounded by inconsistencies in dataset partitioning and
underutilization of advanced LLMs. To address these challenges, we introduce
Similarity-based Neighbor Selection (SNS). Using SimCSE and advanced neighbor
selection techniques, SNS effectively improves the quality of selected
neighbors, thereby improving graph representation and alleviating issues like
over-squashing and heterophily. Besides, as an inductive and training-free
approach, SNS demonstrates superior generalization and scalability over
traditional GNN methods. Our comprehensive experiments, adhering to standard
dataset partitioning practices, demonstrate that SNS, through simple prompt
interactions with LLMs, consistently outperforms vanilla GNNs and achieves
state-of-the-art results on datasets like PubMed in node classification,
showcasing LLMs' potential in graph structure understanding. Our research
further underscores the significance of graph structure integration in LLM
applications and identifies key factors for their success in node
classification. Code is available at https://github.com/ruili33/SNS. | cs.LG | Similarity-based Neighbor Selection for Graph LLMs |
2024-02-06T05:24:16Z | http://arxiv.org/pdf/2402.03719v1 | http://arxiv.org/abs/2402.03719v1 | Jing-Cheng Pang, Heng-Bo Fan, Pengyuan Wang, Jia-Hao Xiao, Nan Tang, Si-Hang Yang, Chengxing Jia, Sheng-Jun Huang, Yang Yu | 2024-02-06T05:24:16Z | cs.CL, cs.AI | The rise of large language models (LLMs) has revolutionized the way that we
interact with artificial intelligence systems through natural language.
However, LLMs often misinterpret user queries because of their uncertain
intention, leading to less helpful responses. In natural human interactions,
clarification is sought through targeted questioning to uncover obscure
information. Thus, in this paper, we introduce LaMAI (Language Model with
Active Inquiry), designed to endow LLMs with this same level of interactive
engagement. LaMAI leverages active learning techniques to raise the most
informative questions, fostering a dynamic bidirectional dialogue. This
approach not only narrows the contextual gap but also refines the output of the
LLMs, aligning it more closely with user expectations. Our empirical studies,
across a variety of complex datasets where LLMs have limited conversational
context, demonstrate the effectiveness of LaMAI. The method improves answer
accuracy from 31.9% to 50.9%, outperforming other leading question-answering
frameworks. Moreover, in scenarios involving human participants, LaMAI
consistently generates responses that are superior or comparable to baseline
methods in more than 82% of the cases. The applicability of LaMAI is further
evidenced by its successful integration with various LLMs, highlighting its
potential for the future of interactive language models. | cs.CL | Empowering Language Models with Active Inquiry for Deeper Understanding |
2024-02-06T05:11:38Z | http://arxiv.org/pdf/2402.03715v1 | http://arxiv.org/abs/2402.03715v1 | Yoonho Lee, Michelle S. Lam, Helena Vasconcelos, Michael S. Bernstein, Chelsea Finn | 2024-02-06T05:11:38Z | cs.LG, cs.AI, cs.CL | In supervised learning, models are trained to extract correlations from a
static dataset. This often leads to models that rely on high-level
misconceptions. To prevent such misconceptions, we must necessarily provide
additional information beyond the training data. Existing methods incorporate
forms of additional instance-level supervision, such as labels for spurious
features or additional labeled data from a balanced distribution. Such
strategies can become prohibitively costly for large-scale datasets since they
require additional annotation at a scale close to the original training data.
We hypothesize that targeted natural language feedback about a model's
misconceptions is a more efficient form of additional supervision. We introduce
Clarify, a novel interface and method for interactively correcting model
misconceptions. Through Clarify, users need only provide a short text
description to describe a model's consistent failure patterns. Then, in an
entirely automated way, we use such descriptions to improve the training
process by reweighting the training data or gathering additional targeted data.
Our user studies show that non-expert users can successfully describe model
misconceptions via Clarify, improving worst-group accuracy by an average of
17.1% in two datasets. Additionally, we use Clarify to find and rectify 31
novel hard subpopulations in the ImageNet dataset, improving minority-split
accuracy from 21.1% to 28.7%. | cs.LG | Clarify: Improving Model Robustness With Natural Language Corrections |
2024-02-06T05:05:38Z | http://arxiv.org/pdf/2402.03710v1 | http://arxiv.org/abs/2402.03710v1 | Xilin Jiang, Cong Han, Yinghao Aaron Li, Nima Mesgarani | 2024-02-06T05:05:38Z | eess.AS, cs.CL, cs.SD | In daily life, we encounter a variety of sounds, both desirable and
undesirable, with limited control over their presence and volume. Our work
introduces "Listen, Chat, and Edit" (LCE), a novel multimodal sound mixture
editor that modifies each sound source in a mixture based on user-provided text
instructions. LCE distinguishes itself with a user-friendly chat interface and
its unique ability to edit multiple sound sources simultaneously within a
mixture, without needing to separate them. Users input open-vocabulary text
prompts, which are interpreted by a large language model to create a semantic
filter for editing the sound mixture. The system then decomposes the mixture
into its components, applies the semantic filter, and reassembles it into the
desired output. We developed a 160-hour dataset with over 100k mixtures,
including speech and various audio sources, along with text prompts for diverse
editing tasks like extraction, removal, and volume control. Our experiments
demonstrate significant improvements in signal quality across all editing tasks
and robust performance in zero-shot scenarios with varying numbers and types of
sound sources. | eess.AS | Listen, Chat, and Edit: Text-Guided Soundscape Modification for Enhanced Auditory Experience |
2024-02-06T04:18:58Z | http://arxiv.org/pdf/2402.05133v1 | http://arxiv.org/abs/2402.05133v1 | Xinyu Li, Zachary C. Lipton, Liu Leqi | 2024-02-06T04:18:58Z | cs.CL, cs.AI, cs.LG | Reinforcement Learning from Human Feedback (RLHF) is the current dominating
framework to fine-tune large language models to better align with human
preferences. However, the underlying premise of algorithms developed under this
framework can be problematic when user preferences encoded in human feedback
are diverse. In this work, we aim to address this problem by developing methods
for building personalized language models. We first formally introduce the task
of learning from personalized human feedback and explain why vanilla RLHF can
be problematic in this context. We then propose a general Personalized-RLHF
(P-RLHF) framework, which requires one to jointly learn a user model and a
language (or reward) model. The user model takes in user information and
outputs user representations. Its structure encodes our assumptions about user
preferences underlying the feedback data. We develop new learning objectives
for personalized reward modeling and personalized Direct Preference
Optimization. To demonstrate the efficacy of our method, we test it on
real-world text summarization data with annotated preferences and annotator
information. We fine-tune GPT-J 6B to obtain personalized language (and reward)
models, which outperform non-personalized models in terms of aligning with
individual preferences. | cs.CL | Personalized Language Modeling from Personalized Human Feedback |
2024-02-06T04:14:09Z | http://arxiv.org/pdf/2402.03686v3 | http://arxiv.org/abs/2402.03686v3 | Soumya Sanyal, Tianyi Xiao, Jiacheng Liu, Wenya Wang, Xiang Ren | 2024-05-27T18:44:14Z | cs.CL, cs.AI | Making inferences in text comprehension to understand the meaning is
essential in language processing. This work studies the entailment verification
(EV) problem of multi-sentence premises that requires a system to make multiple
inferences implicitly. Studying EV for such complex premises is important
because modern NLP problems, such as detecting inconsistent model-generated
rationales, require complex multi-hop reasoning. However, current textual
inference datasets mostly contain short premises that only partially focus on
these challenges. To address this, we compile an EV benchmark that includes
datasets from three NLP domains (NLI, contextual QA, and rationales) containing
multi-sentence premises. On benchmarking humans and LLMs, we find that LLMs are
better than humans in multi-hop reasoning across extended contexts, while
humans perform better in simple deductive reasoning tasks. We also finetune a
Flan-T5 model for EV using two training objectives to obtain a strong
open-source model that outperforms GPT-3.5 and rivals GPT-4. Finally, we use
this model to filter out inconsistent model-generated rationales in
self-consistency decoding, resulting in a 6% accuracy improvement on average
across three MCQ datasets. | cs.CL | Are Machines Better at Complex Reasoning? Unveiling Human-Machine Inference Gaps in Entailment Verification |
2024-02-06T03:41:12Z | http://arxiv.org/pdf/2402.03667v1 | http://arxiv.org/abs/2402.03667v1 | Yanfang Zhang, Yiliu Sun, Yibing Zhan, Dapeng Tao, Dacheng Tao, Chen Gong | 2024-02-06T03:41:12Z | cs.CL, cs.AI | Recently, increasing attention has been focused drawn on to improve the
ability of Large Language Models (LLMs) to perform complex reasoning. However,
previous methods, such as Chain-of-Thought and Self-Consistency, mainly follow
Direct Reasoning (DR) frameworks, so they will meet difficulty in solving
numerous real-world tasks which can hardly be solved via DR. Therefore, to
strengthen the reasoning power of LLMs, this paper proposes a novel Indirect
Reasoning (IR) method that employs the logic of contrapositives and
contradictions to tackle IR tasks such as factual reasoning and mathematic
proof. Specifically, our methodology comprises two steps. Firstly, we leverage
the logical equivalence of contrapositive to augment the data and rules to
enhance the comprehensibility of LLMs. Secondly, we design a set of prompt
templates to trigger LLMs to conduct IR based on proof by contradiction that is
logically equivalent to the original DR process. Our IR method is simple yet
effective and can be straightforwardly integrated with existing DR methods to
further boost the reasoning abilities of LLMs. The experimental results on
popular LLMs, such as GPT-3.5-turbo and Gemini-pro, show that our IR method
enhances the overall accuracy of factual reasoning by 27.33% and mathematical
proof by 31.43%, when compared with traditional DR methods. Moreover, the
methods combining IR and DR significantly outperform the methods solely using
IR or DR, further demonstrating the effectiveness of our strategy. | cs.CL | Large Language Models as an Indirect Reasoner: Contrapositive and Contradiction for Automated Reasoning |
2024-02-06T03:24:27Z | http://arxiv.org/pdf/2402.06049v1 | http://arxiv.org/abs/2402.06049v1 | James Flamino, Mohammed Shahid Modi, Boleslaw K. Szymanski, Brendan Cross, Colton Mikolajczyk | 2024-02-06T03:24:27Z | cs.AI, cs.CL, cs.HC, stat.AP | Large Language Models (LLMs) have shown remarkable promise in their ability
to interact proficiently with humans. Subsequently, their potential use as
artificial confederates and surrogates in sociological experiments involving
conversation is an exciting prospect. But how viable is this idea? This paper
endeavors to test the limits of current-day LLMs with a pre-registered study
integrating real people with LLM agents acting as people. The study focuses on
debate-based opinion consensus formation in three environments: humans only,
agents and humans, and agents only. Our goal is to understand how LLM agents
influence humans, and how capable they are in debating like humans. We find
that LLMs can blend in and facilitate human productivity but are less
convincing in debate, with their behavior ultimately deviating from human's. We
elucidate these primary failings and anticipate that LLMs must evolve further
before being viable debaters. | cs.AI | Limits of Large Language Models in Debating Humans |
2024-02-06T03:18:58Z | http://arxiv.org/pdf/2402.03659v3 | http://arxiv.org/abs/2402.03659v3 | Kelvin J. L. Koa, Yunshan Ma, Ritchie Ng, Tat-Seng Chua | 2024-02-29T12:10:37Z | cs.LG, cs.CL, q-fin.ST | Explaining stock predictions is generally a difficult task for traditional
non-generative deep learning models, where explanations are limited to
visualizing the attention weights on important texts. Today, Large Language
Models (LLMs) present a solution to this problem, given their known
capabilities to generate human-readable explanations for their decision-making
process. However, the task of stock prediction remains challenging for LLMs, as
it requires the ability to weigh the varying impacts of chaotic social texts on
stock prices. The problem gets progressively harder with the introduction of
the explanation component, which requires LLMs to explain verbally why certain
factors are more important than the others. On the other hand, to fine-tune
LLMs for such a task, one would need expert-annotated samples of explanation
for every stock movement in the training set, which is expensive and
impractical to scale. To tackle these issues, we propose our
Summarize-Explain-Predict (SEP) framework, which utilizes a self-reflective
agent and Proximal Policy Optimization (PPO) to let a LLM teach itself how to
generate explainable stock predictions in a fully autonomous manner. The
reflective agent learns how to explain past stock movements through
self-reasoning, while the PPO trainer trains the model to generate the most
likely explanations from input texts. The training samples for the PPO trainer
are also the responses generated during the reflective process, which
eliminates the need for human annotators. Using our SEP framework, we fine-tune
a LLM that can outperform both traditional deep-learning and LLM methods in
prediction accuracy and Matthews correlation coefficient for the stock
classification task. To justify the generalization capability of our framework,
we further test it on the portfolio construction task, and demonstrate its
effectiveness through various portfolio metrics. | cs.LG | Learning to Generate Explainable Stock Predictions using Self-Reflective Large Language Models |
2024-02-06T03:14:46Z | http://arxiv.org/pdf/2402.03658v1 | http://arxiv.org/abs/2402.03658v1 | Kun Ouyang, Liqiang Jing, Xuemeng Song, Meng Liu, Yupeng Hu, Liqiang Nie | 2024-02-06T03:14:46Z | cs.CL, cs.MM | Sarcasm Explanation in Dialogue (SED) is a new yet challenging task, which
aims to generate a natural language explanation for the given sarcastic
dialogue that involves multiple modalities (i.e., utterance, video, and audio).
Although existing studies have achieved great success based on the generative
pretrained language model BART, they overlook exploiting the sentiments
residing in the utterance, video and audio, which are vital clues for sarcasm
explanation. In fact, it is non-trivial to incorporate sentiments for boosting
SED performance, due to three main challenges: 1) diverse effects of utterance
tokens on sentiments; 2) gap between video-audio sentiment signals and the
embedding space of BART; and 3) various relations among utterances, utterance
sentiments, and video-audio sentiments. To tackle these challenges, we propose
a novel sEntiment-enhanceD Graph-based multimodal sarcasm Explanation
framework, named EDGE. In particular, we first propose a lexicon-guided
utterance sentiment inference module, where a heuristic utterance sentiment
refinement strategy is devised. We then develop a module named Joint Cross
Attention-based Sentiment Inference (JCA-SI) by extending the multimodal
sentiment analysis model JCA to derive the joint sentiment label for each
video-audio clip. Thereafter, we devise a context-sentiment graph to
comprehensively model the semantic relations among the utterances, utterance
sentiments, and video-audio sentiments, to facilitate sarcasm explanation
generation. Extensive experiments on the publicly released dataset WITS verify
the superiority of our model over cutting-edge methods. | cs.CL | Sentiment-enhanced Graph-based Sarcasm Explanation in Dialogue |
2024-02-06T02:39:59Z | http://arxiv.org/pdf/2402.03642v1 | http://arxiv.org/abs/2402.03642v1 | Anton Lavrouk, Ian Ligon, Tarek Naous, Jonathan Zheng, Alan Ritter, Wei Xu | 2024-02-06T02:39:59Z | cs.CL, cs.CY, cs.LG, cs.SI | The Stanceosaurus corpus (Zheng et al., 2022) was designed to provide
high-quality, annotated, 5-way stance data extracted from Twitter, suitable for
analyzing cross-cultural and cross-lingual misinformation. In the Stanceosaurus
2.0 iteration, we extend this framework to encompass Russian and Spanish. The
former is of current significance due to prevalent misinformation amid
escalating tensions with the West and the violent incursion into Ukraine. The
latter, meanwhile, represents an enormous community that has been largely
overlooked on major social media platforms. By incorporating an additional
3,874 Spanish and Russian tweets over 41 misinformation claims, our objective
is to support research focused on these issues. To demonstrate the value of
this data, we employed zero-shot cross-lingual transfer on multilingual BERT,
yielding results on par with the initial Stanceosaurus study with a macro F1
score of 43 for both languages. This underlines the viability of stance
classification as an effective tool for identifying multicultural
misinformation. | cs.CL | Stanceosaurus 2.0: Classifying Stance Towards Russian and Spanish Misinformation |
2024-02-06T01:48:53Z | http://arxiv.org/pdf/2402.03628v1 | http://arxiv.org/abs/2402.03628v1 | Zhixuan Chu, Yan Wang, Feng Zhu, Lu Yu, Longfei Li, Jinjie Gu | 2024-02-06T01:48:53Z | cs.CL | The advent of large language models (LLMs) such as ChatGPT, PaLM, and GPT-4
has catalyzed remarkable advances in natural language processing, demonstrating
human-like language fluency and reasoning capacities. This position paper
introduces the concept of Professional Agents (PAgents), an application
framework harnessing LLM capabilities to create autonomous agents with
controllable, specialized, interactive, and professional-level competencies. We
posit that PAgents can reshape professional services through continuously
developed expertise. Our proposed PAgents framework entails a tri-layered
architecture for genesis, evolution, and synergy: a base tool layer, a middle
agent layer, and a top synergy layer. This paper aims to spur discourse on
promising real-world applications of LLMs. We argue the increasing
sophistication and integration of PAgents could lead to AI systems exhibiting
professional mastery over complex domains, serving critical needs, and
potentially achieving artificial general intelligence. | cs.CL | Professional Agents -- Evolving Large Language Models into Autonomous Experts with Human-Level Competencies |
2024-02-06T01:44:38Z | http://arxiv.org/pdf/2402.03627v1 | http://arxiv.org/abs/2402.03627v1 | Hao Wang, Xin Zhang, Jinzhe Jiang, Yaqian Zhao, Chen Li | 2024-02-06T01:44:38Z | cs.CL, cs.AI, I.2.7 | As Large Language Models make a breakthrough in natural language processing
tasks (NLP), multimodal technique becomes extremely popular. However, it has
been shown that multimodal NLP are vulnerable to adversarial attacks, where the
outputs of a model can be dramatically changed by a perturbation to the input.
While several defense techniques have been proposed both in computer vision and
NLP models, the multimodal robustness of models have not been fully explored.
In this paper, we study the adversarial robustness provided by modifying loss
function of pre-trained multimodal models, by restricting top K softmax
outputs. Based on the evaluation and scoring, our experiments show that after a
fine-tuning, adversarial robustness of pre-trained models can be significantly
improved, against popular attacks. Further research should be studying, such as
output diversity, generalization and the robustness-performance trade-off of
this kind of loss functions. Our code will be available after this paper is
accepted | cs.CL | Partially Recentralization Softmax Loss for Vision-Language Models Robustness |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.