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
|
|---|---|---|---|---|---|---|---|---|
2024-02-26T18:59:28Z | http://arxiv.org/pdf/2402.16844v1 | http://arxiv.org/abs/2402.16844v1 | Benjamin Bergner, Andrii Skliar, Amelie Royer, Tijmen Blankevoort, Yuki Asano, Babak Ehteshami Bejnordi | 2024-02-26T18:59:28Z | cs.LG, cs.AI, cs.CL | Large language models (LLMs) have become ubiquitous in practice and are
widely used for generation tasks such as translation, summarization and
instruction following. However, their enormous size and reliance on
autoregressive decoding increase deployment costs and complicate their use in
latency-critical applications. In this work, we propose a hybrid approach that
combines language models of different sizes to increase the efficiency of
autoregressive decoding while maintaining high performance. Our method utilizes
a pretrained frozen LLM that encodes all prompt tokens once in parallel, and
uses the resulting representations to condition and guide a small language
model (SLM), which then generates the response more efficiently. We investigate
the combination of encoder-decoder LLMs with both encoder-decoder and
decoder-only SLMs from different model families and only require fine-tuning of
the SLM. Experiments with various benchmarks show substantial speedups of up to
$4\times$, with minor performance penalties of $1-2\%$ for translation and
summarization tasks compared to the LLM. | cs.LG | Think Big, Generate Quick: LLM-to-SLM for Fast Autoregressive Decoding |
2024-02-26T18:59:18Z | http://arxiv.org/pdf/2402.16843v1 | http://arxiv.org/abs/2402.16843v1 | Ming Zhong, Yelong Shen, Shuohang Wang, Yadong Lu, Yizhu Jiao, Siru Ouyang, Donghan Yu, Jiawei Han, Weizhu Chen | 2024-02-26T18:59:18Z | cs.CV, cs.AI, cs.CL, cs.GR, cs.LG | Low-Rank Adaptation (LoRA) is extensively utilized in text-to-image models
for the accurate rendition of specific elements like distinct characters or
unique styles in generated images. Nonetheless, existing methods face
challenges in effectively composing multiple LoRAs, especially as the number of
LoRAs to be integrated grows, thus hindering the creation of complex imagery.
In this paper, we study multi-LoRA composition through a decoding-centric
perspective. We present two training-free methods: LoRA Switch, which
alternates between different LoRAs at each denoising step, and LoRA Composite,
which simultaneously incorporates all LoRAs to guide more cohesive image
synthesis. To evaluate the proposed approaches, we establish ComposLoRA, a new
comprehensive testbed as part of this research. It features a diverse range of
LoRA categories with 480 composition sets. Utilizing an evaluation framework
based on GPT-4V, our findings demonstrate a clear improvement in performance
with our methods over the prevalent baseline, particularly evident when
increasing the number of LoRAs in a composition. | cs.CV | Multi-LoRA Composition for Image Generation |
2024-02-26T18:59:03Z | http://arxiv.org/pdf/2402.16840v1 | http://arxiv.org/abs/2402.16840v1 | Omkar Thawakar, Ashmal Vayani, Salman Khan, Hisham Cholakal, Rao M. Anwer, Michael Felsberg, Tim Baldwin, Eric P. Xing, Fahad Shahbaz Khan | 2024-02-26T18:59:03Z | cs.CL | "Bigger the better" has been the predominant trend in recent Large Language
Models (LLMs) development. However, LLMs do not suit well for scenarios that
require on-device processing, energy efficiency, low memory footprint, and
response efficiency. These requisites are crucial for privacy, security, and
sustainable deployment. This paper explores the "less is more" paradigm by
addressing the challenge of designing accurate yet efficient Small Language
Models (SLMs) for resource constrained devices. Our primary contribution is the
introduction of an accurate and fully transparent open-source 0.5 billion
(0.5B) parameter SLM, named MobiLlama, catering to the specific needs of
resource-constrained computing with an emphasis on enhanced performance with
reduced resource demands. MobiLlama is a SLM design that initiates from a
larger model and applies a careful parameter sharing scheme to reduce both the
pre-training and the deployment cost. Our work strives to not only bridge the
gap in open-source SLMs but also ensures full transparency, where complete
training data pipeline, training code, model weights, and over 300 checkpoints
along with evaluation codes is available at :
https://github.com/mbzuai-oryx/MobiLlama. | cs.CL | MobiLlama: Towards Accurate and Lightweight Fully Transparent GPT |
2024-02-26T18:57:54Z | http://arxiv.org/pdf/2402.16837v1 | http://arxiv.org/abs/2402.16837v1 | Sohee Yang, Elena Gribovskaya, Nora Kassner, Mor Geva, Sebastian Riedel | 2024-02-26T18:57:54Z | cs.CL | We study whether Large Language Models (LLMs) latently perform multi-hop
reasoning with complex prompts such as "The mother of the singer of
'Superstition' is". We look for evidence of a latent reasoning pathway where an
LLM (1) latently identifies "the singer of 'Superstition'" as Stevie Wonder,
the bridge entity, and (2) uses its knowledge of Stevie Wonder's mother to
complete the prompt. We analyze these two hops individually and consider their
co-occurrence as indicative of latent multi-hop reasoning. For the first hop,
we test if changing the prompt to indirectly mention the bridge entity instead
of any other entity increases the LLM's internal recall of the bridge entity.
For the second hop, we test if increasing this recall causes the LLM to better
utilize what it knows about the bridge entity. We find strong evidence of
latent multi-hop reasoning for the prompts of certain relation types, with the
reasoning pathway used in more than 80% of the prompts. However, the
utilization is highly contextual, varying across different types of prompts.
Also, on average, the evidence for the second hop and the full multi-hop
traversal is rather moderate and only substantial for the first hop. Moreover,
we find a clear scaling trend with increasing model size for the first hop of
reasoning but not for the second hop. Our experimental findings suggest
potential challenges and opportunities for future development and applications
of LLMs. | cs.CL | Do Large Language Models Latently Perform Multi-Hop Reasoning? |
2024-02-26T18:57:52Z | http://arxiv.org/pdf/2402.16836v1 | http://arxiv.org/abs/2402.16836v1 | Dingkun Guo, Yuqi Xiang, Shuqi Zhao, Xinghao Zhu, Masayoshi Tomizuka, Mingyu Ding, Wei Zhan | 2024-02-26T18:57:52Z | cs.RO, cs.AI, cs.CL, cs.CV | Robotic grasping is a fundamental aspect of robot functionality, defining how
robots interact with objects. Despite substantial progress, its
generalizability to counter-intuitive or long-tailed scenarios, such as objects
with uncommon materials or shapes, remains a challenge. In contrast, humans can
easily apply their intuitive physics to grasp skillfully and change grasps
efficiently, even for objects they have never seen before.
This work delves into infusing such physical commonsense reasoning into
robotic manipulation. We introduce PhyGrasp, a multimodal large model that
leverages inputs from two modalities: natural language and 3D point clouds,
seamlessly integrated through a bridge module. The language modality exhibits
robust reasoning capabilities concerning the impacts of diverse physical
properties on grasping, while the 3D modality comprehends object shapes and
parts. With these two capabilities, PhyGrasp is able to accurately assess the
physical properties of object parts and determine optimal grasping poses.
Additionally, the model's language comprehension enables human instruction
interpretation, generating grasping poses that align with human preferences. To
train PhyGrasp, we construct a dataset PhyPartNet with 195K object instances
with varying physical properties and human preferences, alongside their
corresponding language descriptions. Extensive experiments conducted in the
simulation and on the real robots demonstrate that PhyGrasp achieves
state-of-the-art performance, particularly in long-tailed cases, e.g., about
10% improvement in success rate over GraspNet. Project page:
https://sites.google.com/view/phygrasp | cs.RO | PhyGrasp: Generalizing Robotic Grasping with Physics-informed Large Multimodal Models |
2024-02-26T18:57:37Z | http://arxiv.org/pdf/2402.16835v1 | http://arxiv.org/abs/2402.16835v1 | Aengus Lynch, Phillip Guo, Aidan Ewart, Stephen Casper, Dylan Hadfield-Menell | 2024-02-26T18:57:37Z | cs.CL | Machine unlearning can be useful for removing harmful capabilities and
memorized text from large language models (LLMs), but there are not yet
standardized methods for rigorously evaluating it. In this paper, we first
survey techniques and limitations of existing unlearning evaluations. Second,
we apply a comprehensive set of tests for the robustness and competitiveness of
unlearning in the "Who's Harry Potter" (WHP) model from Eldan and Russinovich
(2023). While WHP's unlearning generalizes well when evaluated with the
"Familiarity" metric from Eldan and Russinovich, we find i)
higher-than-baseline amounts of knowledge can reliably be extracted, ii) WHP
performs on par with the original model on Harry Potter Q&A tasks, iii) it
represents latent knowledge comparably to the original model, and iv) there is
collateral unlearning in related domains. Overall, our results highlight the
importance of comprehensive unlearning evaluation that avoids ad-hoc metrics. | cs.CL | Eight Methods to Evaluate Robust Unlearning in LLMs |
2024-02-26T18:56:48Z | http://arxiv.org/pdf/2402.16832v1 | http://arxiv.org/abs/2402.16832v1 | Gaurav Verma, Minje Choi, Kartik Sharma, Jamelle Watson-Daniels, Sejoon Oh, Srijan Kumar | 2024-02-26T18:56:48Z | cs.CL, cs.AI, cs.CV | Multimodal large language models (MLLMs) like LLaVA and GPT-4(V) enable
general-purpose conversations about images with the language modality. As
off-the-shelf MLLMs may have limited capabilities on images from domains like
dermatology and agriculture, they must be fine-tuned to unlock domain-specific
applications. The prevalent architecture of current open-source MLLMs comprises
two major modules: an image-language (cross-modal) projection network and a
large language model. It is desirable to understand the roles of these two
modules in modeling domain-specific visual attributes to inform the design of
future models and streamline the interpretability efforts on the current
models. To this end, via experiments on 4 datasets and under 2 fine-tuning
settings, we find that as the MLLM is fine-tuned, it indeed gains
domain-specific visual capabilities, but the updates do not lead to the
projection extracting relevant domain-specific visual attributes. Our results
indicate that the domain-specific visual attributes are modeled by the LLM,
even when only the projection is fine-tuned. Through this study, we offer a
potential reinterpretation of the role of cross-modal projections in MLLM
architectures. Projection webpage:
https://claws-lab.github.io/projection-in-MLLMs/ | cs.CL | Mysterious Projections: Multimodal LLMs Gain Domain-Specific Visual Capabilities Without Richer Cross-Modal Projections |
2024-02-26T18:56:42Z | http://arxiv.org/pdf/2402.16830v1 | http://arxiv.org/abs/2402.16830v1 | Luca Zampierin, Ghouthi Boukli Hacene, Bac Nguyen, Mirco Ravanelli | 2024-02-26T18:56:42Z | eess.AS, cs.CL, cs.LG, cs.SD | Self-supervised learning (SSL) has achieved remarkable success across various
speech-processing tasks. To enhance its efficiency, previous works often
leverage the use of compression techniques. A notable recent attempt is
DPHuBERT, which applies joint knowledge distillation (KD) and structured
pruning to learn a significantly smaller SSL model. In this paper, we
contribute to this research domain by introducing SKILL, a novel method that
conducts distillation across groups of layers instead of distilling individual
arbitrarily selected layers within the teacher network. The identification of
the layers to distill is achieved through a hierarchical clustering procedure
applied to layer similarity measures. Extensive experiments demonstrate that
our distilled version of WavLM Base+ not only outperforms DPHuBERT but also
achieves state-of-the-art results in the 30M parameters model class across
several SUPERB tasks. | eess.AS | SKILL: Similarity-aware Knowledge distILLation for Speech Self-Supervised Learning |
2024-02-26T18:55:15Z | http://arxiv.org/pdf/2402.16829v1 | http://arxiv.org/abs/2402.16829v1 | Aivin V. Solatorio | 2024-02-26T18:55:15Z | cs.LG, cs.CL | Embedding models are integral to AI applications like semantic search,
personalized recommendations, and retrieval augmented generation for LLMs,
necessitating high-quality training data. However, the limited scalability of
manual data curation prompts the need for automated methods to ensure data
integrity. Traditional unsupervised triplet mining automates training data
generation, crucial for embedding model training, yet inadvertently injects
biases and noise, thereby degrading model performance. Addressing this, we
introduce GISTEmbed, a novel strategy that enhances in-batch negative selection
during contrastive training through a guide model. This approach departs from
reliance on random sampling and equal utility assumption of batch negatives,
significantly reducing noise from data quality issues and improving model
fine-tuning. Benchmarked against the Massive Text Embedding Benchmark (MTEB),
GISTEmbed showcases consistent performance improvements across various model
sizes and achieves state-of-the-art results in select categories. This
framework enables significant enhancements for smaller models by leveraging the
capabilities of powerful yet resource-intensive large models. GISTEmbed can
potentially revolutionize the creation of highly efficient, smaller models,
democratizing access to advanced AI technologies. Making these technologies
more accessible and cost-effective, especially for applications constrained by
resources, significantly expands the impact and accessibility of
state-of-the-art AI solutions across diverse sectors. | cs.LG | GISTEmbed: Guided In-sample Selection of Training Negatives for Text Embedding Fine-tuning |
2024-02-26T18:54:35Z | http://arxiv.org/pdf/2402.16827v2 | http://arxiv.org/abs/2402.16827v2 | Alon Albalak, Yanai Elazar, Sang Michael Xie, Shayne Longpre, Nathan Lambert, Xinyi Wang, Niklas Muennighoff, Bairu Hou, Liangming Pan, Haewon Jeong, Colin Raffel, Shiyu Chang, Tatsunori Hashimoto, William Yang Wang | 2024-03-08T20:04:01Z | cs.CL, cs.LG | A major factor in the recent success of large language models is the use of
enormous and ever-growing text datasets for unsupervised pre-training. However,
naively training a model on all available data may not be optimal (or
feasible), as the quality of available text data can vary. Filtering out data
can also decrease the carbon footprint and financial costs of training models
by reducing the amount of training required. Data selection methods aim to
determine which candidate data points to include in the training dataset and
how to appropriately sample from the selected data points. The promise of
improved data selection methods has caused the volume of research in the area
to rapidly expand. However, because deep learning is mostly driven by empirical
evidence and experimentation on large-scale data is expensive, few
organizations have the resources for extensive data selection research.
Consequently, knowledge of effective data selection practices has become
concentrated within a few organizations, many of which do not openly share
their findings and methodologies. To narrow this gap in knowledge, we present a
comprehensive review of existing literature on data selection methods and
related research areas, providing a taxonomy of existing approaches. By
describing the current landscape of research, this work aims to accelerate
progress in data selection by establishing an entry point for new and
established researchers. Additionally, throughout this review we draw attention
to noticeable holes in the literature and conclude the paper by proposing
promising avenues for future research. | cs.CL | A Survey on Data Selection for Language Models |
2024-02-26T18:48:27Z | http://arxiv.org/pdf/2402.16823v2 | http://arxiv.org/abs/2402.16823v2 | Mingchen Zhuge, Wenyi Wang, Louis Kirsch, Francesco Faccio, Dmitrii Khizbullin, Jürgen Schmidhuber | 2024-02-27T11:03:10Z | cs.AI, cs.CL, cs.LG, cs.MA | Various human-designed prompt engineering techniques have been proposed to
improve problem solvers based on Large Language Models (LLMs), yielding many
disparate code bases. We unify these approaches by describing LLM-based agents
as computational graphs. The nodes implement functions to process multimodal
data or query LLMs, and the edges describe the information flow between
operations. Graphs can be recursively combined into larger composite graphs
representing hierarchies of inter-agent collaboration (where edges connect
operations of different agents). Our novel automatic graph optimizers (1)
refine node-level LLM prompts (node optimization) and (2) improve agent
orchestration by changing graph connectivity (edge optimization). Experiments
demonstrate that our framework can be used to efficiently develop, integrate,
and automatically improve various LLM agents. The code can be found at
https://github.com/metauto-ai/gptswarm. | cs.AI | Language Agents as Optimizable Graphs |
2024-02-26T18:47:27Z | http://arxiv.org/pdf/2402.16822v1 | http://arxiv.org/abs/2402.16822v1 | Mikayel Samvelyan, Sharath Chandra Raparthy, Andrei Lupu, Eric Hambro, Aram H. Markosyan, Manish Bhatt, Yuning Mao, Minqi Jiang, Jack Parker-Holder, Jakob Foerster, Tim Rocktäschel, Roberta Raileanu | 2024-02-26T18:47:27Z | cs.CL, cs.AI, cs.LG | As large language models (LLMs) become increasingly prevalent across many
real-world applications, understanding and enhancing their robustness to user
inputs is of paramount importance. Existing methods for identifying adversarial
prompts tend to focus on specific domains, lack diversity, or require extensive
human annotations. To address these limitations, we present Rainbow Teaming, a
novel approach for producing a diverse collection of adversarial prompts.
Rainbow Teaming casts adversarial prompt generation as a quality-diversity
problem, and uses open-ended search to generate prompts that are both effective
and diverse. It can uncover a model's vulnerabilities across a broad range of
domains including, in this paper, safety, question answering, and
cybersecurity. We also demonstrate that fine-tuning on synthetic data generated
by Rainbow Teaming improves the safety of state-of-the-art LLMs without hurting
their general capabilities and helpfulness, paving the path to open-ended
self-improvement. | cs.CL | Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Prompts |
2024-02-26T18:43:45Z | http://arxiv.org/pdf/2402.16819v2 | http://arxiv.org/abs/2402.16819v2 | Jupinder Parmar, Shrimai Prabhumoye, Joseph Jennings, Mostofa Patwary, Sandeep Subramanian, Dan Su, Chen Zhu, Deepak Narayanan, Aastha Jhunjhunwala, Ayush Dattagupta, Vibhu Jawa, Jiwei Liu, Ameya Mahabaleshwarkar, Osvald Nitski, Annika Brundyn, James Maki, Miguel Martinez, Jiaxuan You, John Kamalu, Patrick LeGresley, Denys Fridman, Jared Casper, Ashwath Aithal, Oleksii Kuchaiev, Mohammad Shoeybi, Jonathan Cohen, Bryan Catanzaro | 2024-02-27T15:22:57Z | cs.CL, cs.AI, cs.LG | We introduce Nemotron-4 15B, a 15-billion-parameter large multilingual
language model trained on 8 trillion text tokens. Nemotron-4 15B demonstrates
strong performance when assessed on English, multilingual, and coding tasks: it
outperforms all existing similarly-sized open models on 4 out of 7 downstream
evaluation areas and achieves competitive performance to the leading open
models in the remaining ones. Specifically, Nemotron-4 15B exhibits the best
multilingual capabilities of all similarly-sized models, even outperforming
models over four times larger and those explicitly specialized for multilingual
tasks. | cs.CL | Nemotron-4 15B Technical Report |
2024-02-26T18:42:26Z | http://arxiv.org/pdf/2402.16817v1 | http://arxiv.org/abs/2402.16817v1 | Jason Phang | 2024-02-26T18:42:26Z | cs.CL | Gisting (Mu et al., 2023) is a simple method for training models to compress
information into fewer token representations using a modified attention mask,
and can serve as an economical approach to training Transformer-based
hypernetworks. We introduce HyperLlama, a set of Gisting-based hypernetworks
built on Llama-2 models that generates task-specific soft prefixes based on
few-shot inputs. In experiments across P3, Super-NaturalInstructions and Symbol
Tuning datasets, we show that HyperLlama models can effectively compress
information from few-shot examples into soft prefixes. However, they still
underperform multi-task fine-tuned language models with full attention over
few-shot in-context examples. We also show that HyperLlama-generated soft
prefixes can serve as better initializations for further prefix tuning.
Overall, Gisting-based hypernetworks are economical and easy to implement, but
have mixed empirical performance. | cs.CL | Investigating the Effectiveness of HyperTuning via Gisting |
2024-02-26T18:37:18Z | http://arxiv.org/pdf/2402.17793v1 | http://arxiv.org/abs/2402.17793v1 | Anne Wu, Kianté Brantley, Yoav Artzi | 2024-02-26T18:37:18Z | cs.AI, cs.CL, cs.LG | This study evaluates three state-of-the-art MLLMs -- GPT-4V, Gemini Pro, and
the open-source model IDEFICS -- on the compositional natural language vision
reasoning task NLVR. Given a human-written sentence paired with a synthetic
image, this task requires the model to determine the truth value of the
sentence with respect to the image. Despite the strong performance demonstrated
by these models, we observe they perform poorly on NLVR, which was constructed
to require compositional and spatial reasoning, and to be robust for semantic
and systematic biases. | cs.AI | A Surprising Failure? Multimodal LLMs and the NLVR Challenge |
2024-02-26T18:33:13Z | http://arxiv.org/pdf/2402.16810v1 | http://arxiv.org/abs/2402.16810v1 | Fujian Jia, Xin Liu, Lixi Deng, Jiwen Gu, Chunchao Pu, Tunan Bai, Mengjiang Huang, Yuanzhi Lu, Kang Liu | 2024-02-26T18:33:13Z | cs.CL | In the past year, there has been a growing trend in applying Large Language
Models (LLMs) to the field of medicine, particularly with the advent of
advanced language models such as ChatGPT developed by OpenAI. However, there is
limited research on LLMs specifically addressing oncology-related queries. The
primary aim of this research was to develop a specialized language model that
demonstrates improved accuracy in providing advice related to oncology. We
performed an extensive data collection of online question-answer interactions
centered around oncology, sourced from reputable doctor-patient platforms.
Following data cleaning and anonymization, a dataset comprising over 180K+
oncology-related conversations was established. The conversations were
categorized and meticulously reviewed by field specialists and clinicians to
ensure precision. Employing the LLaMA model and other selected open-source
datasets, we conducted iterative fine-tuning to enhance the model's proficiency
in basic medical conversation and specialized oncology knowledge. We observed a
substantial enhancement in the model's understanding of genuine patient
inquiries and its reliability in offering oncology-related advice through the
utilization of real online question-answer interactions in the fine-tuning
process. We release database and models to the research community
(https://github.com/OncoGPT1). | cs.CL | OncoGPT: A Medical Conversational Model Tailored with Oncology Domain Expertise on a Large Language Model Meta-AI (LLaMA) |
2024-02-26T18:10:56Z | http://arxiv.org/pdf/2402.16797v2 | http://arxiv.org/abs/2402.16797v2 | Bowen Zhao, Zander Brumbaugh, Yizhong Wang, Hannaneh Hajishirzi, Noah A. Smith | 2024-06-09T19:47:15Z | cs.CL | Language models (LMs) are trained on web text originating from many points in
time and, in general, without any explicit temporal grounding. This work
investigates the temporal chaos of pretrained LMs and explores various methods
to align their internal knowledge to a target time, which we call "temporal
alignment." To do this, we first automatically construct a dataset containing
20K time-sensitive questions and their answers for each year from 2000 to 2023.
Based on this dataset, we empirically show that pretrained LMs (e.g., LLaMa2),
despite having a recent pretraining cutoff (e.g., 2022), mostly answer
questions using earlier knowledge (e.g., in 2019). We then develop several
methods, from prompting to finetuning, to align LMs to use their most recent
knowledge when answering questions, and investigate various factors in this
alignment. Our experiments demonstrate that aligning LLaMa2 to the year 2022
can enhance its performance by up to 62% according to that year's answers. This
improvement occurs even without explicitly mentioning time information,
indicating the possibility of aligning models' internal sense of time after
pretraining. Finally, we find that alignment to a historical time is also
possible, with up to 2.8$\times$ the performance of the unaligned LM in 2010 if
finetuning models to that year. These findings hint at the sophistication of
LMs' internal knowledge organization and the necessity of tuning them properly. | cs.CL | Set the Clock: Temporal Alignment of Pretrained Language Models |
2024-02-26T18:08:52Z | http://arxiv.org/pdf/2402.16795v2 | http://arxiv.org/abs/2402.16795v2 | Zeyu He, Chieh-Yang Huang, Chien-Kuang Cornelia Ding, Shaurya Rohatgi, Ting-Hao 'Kenneth' Huang | 2024-06-28T19:33:48Z | cs.HC, cs.AI, cs.CL, cs.LG | Recent studies indicated GPT-4 outperforms online crowd workers in data
labeling accuracy, notably workers from Amazon Mechanical Turk (MTurk).
However, these studies were criticized for deviating from standard
crowdsourcing practices and emphasizing individual workers' performances over
the whole data-annotation process. This paper compared GPT-4 and an ethical and
well-executed MTurk pipeline, with 415 workers labeling 3,177 sentence segments
from 200 scholarly articles using the CODA-19 scheme. Two worker interfaces
yielded 127,080 labels, which were then used to infer the final labels through
eight label-aggregation algorithms. Our evaluation showed that despite best
practices, MTurk pipeline's highest accuracy was 81.5%, whereas GPT-4 achieved
83.6%. Interestingly, when combining GPT-4's labels with crowd labels collected
via an advanced worker interface for aggregation, 2 out of the 8 algorithms
achieved an even higher accuracy (87.5%, 87.0%). Further analysis suggested
that, when the crowd's and GPT-4's labeling strengths are complementary,
aggregating them could increase labeling accuracy. | cs.HC | If in a Crowdsourced Data Annotation Pipeline, a GPT-4 |
2024-02-26T18:00:49Z | http://arxiv.org/pdf/2402.16786v2 | http://arxiv.org/abs/2402.16786v2 | Paul Röttger, Valentin Hofmann, Valentina Pyatkin, Musashi Hinck, Hannah Rose Kirk, Hinrich Schütze, Dirk Hovy | 2024-06-05T10:17:53Z | cs.CL, cs.AI | Much recent work seeks to evaluate values and opinions in large language
models (LLMs) using multiple-choice surveys and questionnaires. Most of this
work is motivated by concerns around real-world LLM applications. For example,
politically-biased LLMs may subtly influence society when they are used by
millions of people. Such real-world concerns, however, stand in stark contrast
to the artificiality of current evaluations: real users do not typically ask
LLMs survey questions. Motivated by this discrepancy, we challenge the
prevailing constrained evaluation paradigm for values and opinions in LLMs and
explore more realistic unconstrained evaluations. As a case study, we focus on
the popular Political Compass Test (PCT). In a systematic review, we find that
most prior work using the PCT forces models to comply with the PCT's
multiple-choice format. We show that models give substantively different
answers when not forced; that answers change depending on how models are
forced; and that answers lack paraphrase robustness. Then, we demonstrate that
models give different answers yet again in a more realistic open-ended answer
setting. We distill these findings into recommendations and open challenges in
evaluating values and opinions in LLMs. | cs.CL | Political Compass or Spinning Arrow? Towards More Meaningful Evaluations for Values and Opinions in Large Language Models |
2024-02-26T17:45:36Z | http://arxiv.org/pdf/2402.16775v2 | http://arxiv.org/abs/2402.16775v2 | Renren Jin, Jiangcun Du, Wuwei Huang, Wei Liu, Jian Luan, Bin Wang, Deyi Xiong | 2024-06-06T13:38:26Z | cs.CL, cs.AI | Increasing the number of parameters in large language models (LLMs) usually
improves performance in downstream tasks but raises compute and memory costs,
making deployment difficult in resource-limited settings. Quantization
techniques, which reduce the bits needed for model weights or activations with
minimal performance loss, have become popular due to the rise of LLMs. However,
most quantization studies use pre-trained LLMs, and the impact of quantization
on instruction-tuned LLMs and the relationship between perplexity and benchmark
performance of quantized LLMs are not well understood. Evaluation of quantized
LLMs is often limited to language modeling and a few classification tasks,
leaving their performance on other benchmarks unclear. To address these gaps,
we propose a structured evaluation framework consisting of three critical
dimensions: (1) knowledge \& capacity, (2) alignment, and (3) efficiency, and
conduct extensive experiments across ten diverse benchmarks. Our experimental
results indicate that LLMs with 4-bit quantization can retain performance
comparable to their non-quantized counterparts, and perplexity can serve as a
proxy metric for quantized LLMs on most benchmarks. Furthermore, quantized LLMs
with larger parameter scales can outperform smaller LLMs. Despite the memory
savings achieved through quantization, it can also slow down the inference
speed of LLMs. Consequently, substantial engineering efforts and hardware
support are imperative to achieve a balanced optimization of decoding speed and
memory consumption in the context of quantized LLMs. | cs.CL | A Comprehensive Evaluation of Quantization Strategies for Large Language Models |
2024-02-26T17:35:44Z | http://arxiv.org/pdf/2402.16767v1 | http://arxiv.org/abs/2402.16767v1 | Jiafeng Guo, Changjiang Zhou, Ruqing Zhang, Jiangui Chen, Maarten de Rijke, Yixing Fan, Xueqi Cheng | 2024-02-26T17:35:44Z | cs.IR, cs.CL | Knowledge-intensive language tasks (KILTs) typically require retrieving
relevant documents from trustworthy corpora, e.g., Wikipedia, to produce
specific answers. Very recently, a pre-trained generative retrieval model for
KILTs, named CorpusBrain, was proposed and reached new state-of-the-art
retrieval performance. However, most existing research on KILTs, including
CorpusBrain, has predominantly focused on a static document collection,
overlooking the dynamic nature of real-world scenarios, where new documents are
continuously being incorporated into the source corpus. To address this gap, it
is crucial to explore the capability of retrieval models to effectively handle
the dynamic retrieval scenario inherent in KILTs.
In this work, we first introduce the continual document learning (CDL) task
for KILTs and build a novel benchmark dataset named KILT++ based on the
original KILT dataset for evaluation. Then, we conduct a comprehensive study
over the use of pre-trained CorpusBrain on KILT++. Unlike the promising results
in the stationary scenario, CorpusBrain is prone to catastrophic forgetting in
the dynamic scenario, hence hampering the retrieval performance. To alleviate
this issue, we propose CorpusBrain++, a continual generative pre-training
framework. Empirical results demonstrate the significant effectiveness and
remarkable efficiency of CorpusBrain++ in comparison to both traditional and
generative IR methods. | cs.IR | CorpusBrain++: A Continual Generative Pre-Training Framework for Knowledge-Intensive Language Tasks |
2024-02-26T17:16:28Z | http://arxiv.org/pdf/2402.16751v1 | http://arxiv.org/abs/2402.16751v1 | Enrico Liscio, Luciano C. Siebert, Catholijn M. Jonker, Pradeep K. Murukannaiah | 2024-02-26T17:16:28Z | cs.AI, cs.CL | Understanding citizens' values in participatory systems is crucial for
citizen-centric policy-making. We envision a hybrid participatory system where
participants make choices and provide motivations for those choices, and AI
agents estimate their value preferences by interacting with them. We focus on
situations where a conflict is detected between participants' choices and
motivations, and propose methods for estimating value preferences while
addressing detected inconsistencies by interacting with the participants. We
operationalize the philosophical stance that "valuing is deliberatively
consequential." That is, if a participant's choice is based on a deliberation
of value preferences, the value preferences can be observed in the motivation
the participant provides for the choice. Thus, we propose and compare value
estimation methods that prioritize the values estimated from motivations over
the values estimated from choices alone. Then, we introduce a disambiguation
strategy that addresses the detected inconsistencies between choices and
motivations by directly interacting with the participants. We evaluate the
proposed methods on a dataset of a large-scale survey on energy transition. The
results show that explicitly addressing inconsistencies between choices and
motivations improves the estimation of an individual's value preferences. The
disambiguation strategy does not show substantial improvements when compared to
similar baselines--however, we discuss how the novelty of the approach can open
new research avenues and propose improvements to address the current
limitations. | cs.AI | Value Preferences Estimation and Disambiguation in Hybrid Participatory Systems |
2024-02-26T16:35:59Z | http://arxiv.org/pdf/2402.16717v1 | http://arxiv.org/abs/2402.16717v1 | Huijie Lv, Xiao Wang, Yuansen Zhang, Caishuang Huang, Shihan Dou, Junjie Ye, Tao Gui, Qi Zhang, Xuanjing Huang | 2024-02-26T16:35:59Z | cs.CL, cs.AI, cs.CR | Adversarial misuse, particularly through `jailbreaking' that circumvents a
model's safety and ethical protocols, poses a significant challenge for Large
Language Models (LLMs). This paper delves into the mechanisms behind such
successful attacks, introducing a hypothesis for the safety mechanism of
aligned LLMs: intent security recognition followed by response generation.
Grounded in this hypothesis, we propose CodeChameleon, a novel jailbreak
framework based on personalized encryption tactics. To elude the intent
security recognition phase, we reformulate tasks into a code completion format,
enabling users to encrypt queries using personalized encryption functions. To
guarantee response generation functionality, we embed a decryption function
within the instructions, which allows the LLM to decrypt and execute the
encrypted queries successfully. We conduct extensive experiments on 7 LLMs,
achieving state-of-the-art average Attack Success Rate (ASR). Remarkably, our
method achieves an 86.6\% ASR on GPT-4-1106. | cs.CL | CodeChameleon: Personalized Encryption Framework for Jailbreaking Large Language Models |
2024-02-26T16:31:28Z | http://arxiv.org/pdf/2402.16714v2 | http://arxiv.org/abs/2402.16714v2 | Naixu Guo, Zhan Yu, Matthew Choi, Aman Agrawal, Kouhei Nakaji, Alán Aspuru-Guzik, Patrick Rebentrost | 2024-05-31T03:34:57Z | quant-ph, cs.AI, cs.CL | Generative machine learning methods such as large-language models are
revolutionizing the creation of text and images. While these models are
powerful they also harness a large amount of computational resources. The
transformer is a key component in large language models that aims to generate a
suitable completion of a given partial sequence. In this work, we investigate
transformer architectures under the lens of fault-tolerant quantum computing.
The input model is one where trained weight matrices are given as block
encodings and we construct the query, key, and value matrices for the
transformer. We show how to prepare a block encoding of the self-attention
matrix, with a new subroutine for the row-wise application of the softmax
function. In addition, we combine quantum subroutines to construct important
building blocks in the transformer, the residual connection and layer
normalization, and the feed-forward neural network. Our subroutines prepare an
amplitude encoding of the transformer output, which can be measured to obtain a
prediction. Based on common open-source large-language models, we provide
insights into the behavior of important parameters determining the run time of
the quantum algorithm. We discuss the potential and challenges for obtaining a
quantum advantage. | quant-ph | Quantum linear algebra is all you need for Transformer architectures |
2024-02-26T16:21:53Z | http://arxiv.org/pdf/2402.16705v1 | http://arxiv.org/abs/2402.16705v1 | Liangxin Liu, Xuebo Liu, Derek F. Wong, Dongfang Li, Ziyi Wang, Baotian Hu, Min Zhang | 2024-02-26T16:21:53Z | cs.CL, cs.AI, cs.LG | Instruction tuning (IT) is crucial to tailoring large language models (LLMs)
towards human-centric interactions. Recent advancements have shown that the
careful selection of a small, high-quality subset of IT data can significantly
enhance the performance of LLMs. Despite this, common approaches often rely on
additional models or data sets, which increases costs and limits widespread
adoption. In this work, we propose a novel approach, termed SelectIT, that
capitalizes on the foundational capabilities of the LLM itself. Specifically,
we exploit the intrinsic uncertainty present in LLMs to more effectively select
high-quality IT data, without the need for extra resources. Furthermore, we
introduce a novel IT dataset, the Selective Alpaca, created by applying
SelectIT to the Alpaca-GPT4 dataset. Empirical results demonstrate that IT
using Selective Alpaca leads to substantial model ability enhancement. The
robustness of SelectIT has also been corroborated in various foundation models
and domain-specific tasks. Our findings suggest that longer and more
computationally intensive IT data may serve as superior sources of IT, offering
valuable insights for future research in this area. Data, code, and scripts are
freely available at https://github.com/Blue-Raincoat/SelectIT. | cs.CL | SelectIT: Selective Instruction Tuning for Large Language Models via Uncertainty-Aware Self-Reflection |
2024-02-26T16:17:19Z | http://arxiv.org/pdf/2403.00821v1 | http://arxiv.org/abs/2403.00821v1 | Seibi Kobara, Alireza Rafiei, Masoud Nateghi, Selen Bozkurt, Rishikesan Kamaleswaran, Abeed Sarker | 2024-02-26T16:17:19Z | cs.CL, cs.LG, cs.SI | Breast cancer is a significant public health concern and is the leading cause
of cancer-related deaths among women. Despite advances in breast cancer
treatments, medication non-adherence remains a major problem. As electronic
health records do not typically capture patient-reported outcomes that may
reveal information about medication-related experiences, social media presents
an attractive resource for enhancing our understanding of the patients'
treatment experiences. In this paper, we developed natural language processing
(NLP) based methodologies to study information posted by an automatically
curated breast cancer cohort from social media. We employed a transformer-based
classifier to identify breast cancer patients/survivors on X (Twitter) based on
their self-reported information, and we collected longitudinal data from their
profiles. We then designed a multi-layer rule-based model to develop a breast
cancer therapy-associated side effect lexicon and detect patterns of medication
usage and associated side effects among breast cancer patients. 1,454,637 posts
were available from 583,962 unique users, of which 62,042 were detected as
breast cancer members using our transformer-based model. 198 cohort members
mentioned breast cancer medications with tamoxifen as the most common. Our side
effect lexicon identified well-known side effects of hormone and chemotherapy.
Furthermore, it discovered a subject feeling towards cancer and medications,
which may suggest a pre-clinical phase of side effects or emotional distress.
This analysis highlighted not only the utility of NLP techniques in
unstructured social media data to identify self-reported breast cancer posts,
medication usage patterns, and treatment side effects but also the richness of
social data on such clinical questions. | cs.CL | Social Media as a Sensor: Analyzing Twitter Data for Breast Cancer Medication Effects Using Natural Language Processing |
2024-02-26T16:14:47Z | http://arxiv.org/pdf/2402.16700v1 | http://arxiv.org/abs/2402.16700v1 | Itay Etelis, Avi Rosenfeld, Abraham Itzhak Weinberg, David Sarne | 2024-02-26T16:14:47Z | cs.CL, cs.AI | In recent years, transformer models have revolutionized Natural Language
Processing (NLP), achieving exceptional results across various tasks, including
Sentiment Analysis (SA). As such, current state-of-the-art approaches for SA
predominantly rely on transformer models alone, achieving impressive accuracy
levels on benchmark datasets. In this paper, we show that the key for further
improving the accuracy of such ensembles for SA is to include not only
transformers, but also traditional NLP models, despite the inferiority of the
latter compared to transformer models. However, as we empirically show, this
necessitates a change in how the ensemble is constructed, specifically relying
on the Hierarchical Ensemble Construction (HEC) algorithm we present. Our
empirical studies across eight canonical SA datasets reveal that ensembles
incorporating a mix of model types, structured via HEC, significantly
outperform traditional ensembles. Finally, we provide a comparative analysis of
the performance of the HEC and GPT-4, demonstrating that while GPT-4 closely
approaches state-of-the-art SA methods, it remains outperformed by our proposed
ensemble strategy. | cs.CL | Generating Effective Ensembles for Sentiment Analysis |
2024-02-26T16:11:03Z | http://arxiv.org/pdf/2402.16696v2 | http://arxiv.org/abs/2402.16696v2 | Anchun Gui, Jian Li, Yong Dai, Nan Du, Han Xiao | 2024-02-28T03:31:28Z | cs.CL | Tool-augmented large language models (LLMs) are attracting widespread
attention when accessing up-to-date knowledge and alleviating hallucination
issues. Nowadays, advanced closed-source LLMs (e.g., ChatGPT) have demonstrated
surprising tool-usage capabilities through prompting and in-context learning
techniques. To empower the capabilities of open-source LLMs (e.g., LLaMA) in
manipulating tools, current efforts focus on either template-driven or
token-triggered tool-usage. However, the former hampers LLMs' flexibility to
address diverse user's queries due to constrained tool interactions, while the
latter limits the generalizability when engaging with new tools, since
tool-usage learning is based on task- and tool-specific datasets. To alleviate
these concerns, in this paper, we propose a decision-aware and generalizable
tool-usage framework (DEER). Specifically, we first construct the tool-usage
samples with multiple decision branches via an automatic generation pipeline,
thereby inspiring the decision-making awareness of LLMs under diverse
scenarios. Meanwhile, we propose a novel tool sampling strategy to enhance the
generalizability of LLMs over unseen tools. Extensive experiments demonstrate
that our proposed DEER is effective and significantly outperforms baselines
across various datasets. | cs.CL | Look Before You Leap: Towards Decision-Aware and Generalizable Tool-Usage for Large Language Models |
2024-02-26T16:09:00Z | http://arxiv.org/pdf/2402.16694v2 | http://arxiv.org/abs/2402.16694v2 | Qiwei Peng, Yekun Chai, Xuhong Li | 2024-03-24T15:41:21Z | cs.CL, cs.PL, cs.SE | Large language models (LLMs) have made significant progress in generating
codes from textual prompts. However, existing benchmarks have mainly
concentrated on translating English prompts to multilingual codes or have been
constrained to very limited natural languages (NLs). These benchmarks have
overlooked the vast landscape of massively multilingual NL to multilingual
code, leaving a critical gap in the evaluation of multilingual LLMs. In
response, we introduce HumanEval-XL, a massively multilingual code generation
benchmark specifically crafted to address this deficiency. HumanEval-XL
establishes connections between 23 NLs and 12 programming languages (PLs), and
comprises of a collection of 22,080 prompts with an average of 8.33 test cases.
By ensuring parallel data across multiple NLs and PLs, HumanEval-XL offers a
comprehensive evaluation platform for multilingual LLMs, allowing the
assessment of the understanding of different NLs. Our work serves as a
pioneering step towards filling the void in evaluating NL generalization in the
area of multilingual code generation. We make our evaluation code and data
publicly available at \url{https://github.com/FloatAI/humaneval-xl}. | cs.CL | HumanEval-XL: A Multilingual Code Generation Benchmark for Cross-lingual Natural Language Generalization |
2024-02-26T16:05:33Z | http://arxiv.org/pdf/2402.16689v1 | http://arxiv.org/abs/2402.16689v1 | Adrien Bazoge, Emmanuel Morin, Beatrice Daille, Pierre-Antoine Gourraud | 2024-02-26T16:05:33Z | cs.CL, cs.AI | Recently, pretrained language models based on BERT have been introduced for
the French biomedical domain. Although these models have achieved
state-of-the-art results on biomedical and clinical NLP tasks, they are
constrained by a limited input sequence length of 512 tokens, which poses
challenges when applied to clinical notes. In this paper, we present a
comparative study of three adaptation strategies for long-sequence models,
leveraging the Longformer architecture. We conducted evaluations of these
models on 16 downstream tasks spanning both biomedical and clinical domains.
Our findings reveal that further pre-training an English clinical model with
French biomedical texts can outperform both converting a French biomedical BERT
to the Longformer architecture and pre-training a French biomedical Longformer
from scratch. The results underscore that long-sequence French biomedical
models improve performance across most downstream tasks regardless of sequence
length, but BERT based models remain the most efficient for named entity
recognition tasks. | cs.CL | Adaptation of Biomedical and Clinical Pretrained Models to French Long Documents: A Comparative Study |
2024-02-26T15:47:01Z | http://arxiv.org/pdf/2402.16671v6 | http://arxiv.org/abs/2402.16671v6 | Alex Zhuang, Ge Zhang, Tianyu Zheng, Xinrun Du, Junjie Wang, Weiming Ren, Stephen W. Huang, Jie Fu, Xiang Yue, Wenhu Chen | 2024-04-24T21:13:24Z | cs.CL | Structured data sources, such as tables, graphs, and databases, are
ubiquitous knowledge sources. Despite the demonstrated capabilities of large
language models (LLMs) on plain text, their proficiency in interpreting and
utilizing structured data remains limited. Our investigation reveals a notable
deficiency in LLMs' ability to process structured data, e.g., ChatGPT lags
behind state-of-the-art (SoTA) model by an average of 35%. To augment the
Structured Knowledge Grounding (SKG) capabilities in LLMs, we have developed a
comprehensive instruction tuning dataset comprising 1.1 million examples.
Utilizing this dataset, we train a series of models, referred to as StructLM,
based on the Mistral and the CodeLlama model family, ranging from 7B to 34B
parameters. Our StructLM series surpasses task-specific models on 16 out of 18
evaluated datasets and establishes new SoTA performance on 8 SKG tasks.
Furthermore, StructLM demonstrates strong generalization across 6 novel
held-out SKG tasks, outperforming TableLlama by an average of 35\% and Flan-UL2
20B by an average of 10\%. Contrary to expectations, we observe that scaling
model size offers marginal benefits, with StructLM-34B showing only slight
improvements over StructLM-7B. This suggests that structured knowledge
grounding is still a challenging task and requires more innovative design to
push to a new level. | cs.CL | StructLM: Towards Building Generalist Models for Structured Knowledge Grounding |
2024-02-26T15:39:52Z | http://arxiv.org/pdf/2402.16667v1 | http://arxiv.org/abs/2402.16667v1 | Qinyu Luo, Yining Ye, Shihao Liang, Zhong Zhang, Yujia Qin, Yaxi Lu, Yesai Wu, Xin Cong, Yankai Lin, Yingli Zhang, Xiaoyin Che, Zhiyuan Liu, Maosong Sun | 2024-02-26T15:39:52Z | cs.CL, cs.AI, I.2.7; F.2.2 | Generative models have demonstrated considerable potential in software
engineering, particularly in tasks such as code generation and debugging.
However, their utilization in the domain of code documentation generation
remains underexplored. To this end, we introduce RepoAgent, a large language
model powered open-source framework aimed at proactively generating,
maintaining, and updating code documentation. Through both qualitative and
quantitative evaluations, we have validated the effectiveness of our approach,
showing that RepoAgent excels in generating high-quality repository-level
documentation. The code and results are publicly accessible at
https://github.com/OpenBMB/RepoAgent. | cs.CL | RepoAgent: An LLM-Powered Open-Source Framework for Repository-level Code Documentation Generation |
2024-02-26T15:26:56Z | http://arxiv.org/pdf/2402.16654v1 | http://arxiv.org/abs/2402.16654v1 | Pavel Blinov, Konstantin Egorov, Ivan Sviridov, Nikolay Ivanov, Stepan Botman, Evgeniy Tagin, Stepan Kudin, Galina Zubkova, Andrey Savchenko | 2024-02-26T15:26:56Z | cs.AI, cs.CL, cs.HC, 68T07, I.2.1 | Building an intelligent and efficient medical assistant is still a
challenging AI problem. The major limitation comes from the data modality
scarceness, which reduces comprehensive patient perception. This demo paper
presents the GigaPevt, the first multimodal medical assistant that combines the
dialog capabilities of large language models with specialized medical models.
Such an approach shows immediate advantages in dialog quality and metric
performance, with a 1.18\% accuracy improvement in the question-answering task. | cs.AI | GigaPevt: Multimodal Medical Assistant |
2024-02-26T15:22:30Z | http://arxiv.org/pdf/2402.16650v1 | http://arxiv.org/abs/2402.16650v1 | Karim Derrick | 2024-02-26T15:22:30Z | cs.CL, cs.CE, cs.CY | In this paper we explore the challenges of measuring sentiment in relation to
Environmental, Social and Governance (ESG) social media. ESG has grown in
importance in recent years with a surge in interest from the financial sector
and the performance of many businesses has become based in part on their ESG
related reputations. The use of sentiment analysis to measure ESG related
reputation has developed and with it interest in the use of machines to do so.
The era of digital media has created an explosion of new media sources, driven
by the growth of social media platforms. This growing data environment has
become an excellent source for behavioural insight studies across many
disciplines that includes politics, healthcare and market research. Our study
seeks to compare human performance with the cutting edge in machine performance
in the measurement of ESG related sentiment. To this end researchers classify
the sentiment of 150 tweets and a reliability measure is made. A gold standard
data set is then established based on the consensus of 3 researchers and this
data set is then used to measure the performance of different machine
approaches: one based on the VADER dictionary approach to sentiment
classification and then multiple language model approaches, including Llama2,
T5, Mistral, Mixtral, FINBERT, GPT3.5 and GPT4. | cs.CL | ESG Sentiment Analysis: comparing human and language model performance including GPT |
2024-02-26T15:05:16Z | http://arxiv.org/pdf/2402.16929v2 | http://arxiv.org/abs/2402.16929v2 | Ming Wang, Yuanzhong Liu, Xiaoyu Liang, Songlian Li, Yijie Huang, Xiaoming Zhang, Sijia Shen, Chaofeng Guan, Daling Wang, Shi Feng, Huaiwen Zhang, Yifei Zhang, Minghui Zheng, Chi Zhang | 2024-06-29T14:19:08Z | cs.SE, cs.AI, cs.CL, cs.PL | LLMs have demonstrated commendable performance across diverse domains.
Nevertheless, formulating high-quality prompts to instruct LLMs proficiently
poses a challenge for non-AI experts. Existing research in prompt engineering
suggests somewhat scattered optimization principles and designs empirically
dependent prompt optimizers. Unfortunately, these endeavors lack a structured
design template, incurring high learning costs and resulting in low
reusability. In addition, it is not conducive to the iterative updating of
prompts. Inspired by structured reusable programming languages, we propose
LangGPT, a dual-layer prompt design framework as the programming language for
LLMs. LangGPT has an easy-to-learn normative structure and provides an extended
structure for migration and reuse. Experiments illustrate that LangGPT
significantly enhances the performance of LLMs. Moreover, the case study shows
that LangGPT leads LLMs to generate higher-quality responses. Furthermore, we
analyzed the ease of use and reusability of LangGPT through a user survey in
our online community. | cs.SE | LangGPT: Rethinking Structured Reusable Prompt Design Framework for LLMs from the Programming Language |
2024-02-26T15:04:35Z | http://arxiv.org/pdf/2402.16632v1 | http://arxiv.org/abs/2402.16632v1 | Alessandro Maisto | 2024-02-26T15:04:35Z | cs.CL | In this work, we propose a Distributional Semantic resource enriched with
linguistic and lexical information extracted from electronic dictionaries,
designed to address the challenge of bridging the gap between the continuous
semantic values represented by distributional vectors and the discrete
descriptions offered by general semantics theory. Recently, many researchers
have concentrated on the nexus between embeddings and a comprehensive theory of
semantics and meaning. This often involves decoding the representation of word
meanings in Distributional Models into a set of discrete, manually constructed
properties such as semantic primitives or features, using neural decoding
techniques. Our approach introduces an alternative strategy grounded in
linguistic data. We have developed a collection of domain-specific
co-occurrence matrices, derived from two sources: a classification of Italian
nouns categorized into 4 semantic traits and 20 concrete noun sub-categories,
and a list of Italian verbs classified according to their semantic classes. In
these matrices, the co-occurrence values for each word are calculated
exclusively with a defined set of words pertinent to a particular lexical
domain. The resource comprises 21 domain-specific matrices, one comprehensive
matrix, and a Graphical User Interface. Our model facilitates the generation of
reasoned semantic descriptions of concepts by selecting matrices directly
associated with concrete conceptual knowledge, such as a matrix based on
location nouns and the concept of animal habitats. We assessed the utility of
the resource through two experiments, achieving promising outcomes in both: the
automatic classification of animal nouns and the extraction of animal features. | cs.CL | Domain Embeddings for Generating Complex Descriptions of Concepts in Italian Language |
2024-02-26T14:47:35Z | http://arxiv.org/pdf/2402.16617v2 | http://arxiv.org/abs/2402.16617v2 | Howard Yen, Tianyu Gao, Danqi Chen | 2024-06-11T18:54:13Z | cs.CL | Extending large language models (LLMs) to process longer inputs is crucial
for a wide range of applications. However, the substantial computational cost
of transformers and limited generalization of positional encoding restrict the
size of their context window. We introduce Context Expansion with Parallel
Encoding (CEPE), a framework that can be applied to any existing decoder-only
LLMs to extend their context window. CEPE employs a small encoder to process
long inputs chunk by chunk, enabling the frozen decoder to utilize additional
contexts via cross-attention. CEPE is efficient, generalizable, and versatile:
trained with 8K-token documents, it extends the context window of LLAMA-2 to
128K tokens, offering 10x the throughput with only 1/6 of the memory. CEPE
yields strong performance on language modeling and in-context learning. CEPE
also excels in retrieval-augmented applications, while existing long-context
models degenerate with retrieved contexts. We further introduce a CEPE variant
that can extend the context window of instruction-tuned models using only
unlabeled data, and showcase its effectiveness on LLAMA-2-CHAT, leading to a
strong instruction-following model that can leverage very long contexts on
downstream tasks. | cs.CL | Long-Context Language Modeling with Parallel Context Encoding |
2024-02-26T14:40:34Z | http://arxiv.org/pdf/2402.16608v1 | http://arxiv.org/abs/2402.16608v1 | Pierre Erbacher, Jian-Yun Nie, Philippe Preux, Laure Soulier | 2024-02-26T14:40:34Z | cs.CL, cs.IR | Conversational systems have made significant progress in generating natural
language responses. However, their potential as conversational search systems
is currently limited due to their passive role in the information-seeking
process. One major limitation is the scarcity of datasets that provide labelled
ambiguous questions along with a supporting corpus of documents and relevant
clarifying questions. This work aims to tackle the challenge of generating
relevant clarifying questions by taking into account the inherent ambiguities
present in both user queries and documents. To achieve this, we propose PAQA,
an extension to the existing AmbiNQ dataset, incorporating clarifying
questions. We then evaluate various models and assess how passage retrieval
impacts ambiguity detection and the generation of clarifying questions. By
addressing this gap in conversational search systems, we aim to provide
additional supervision to enhance their active participation in the
information-seeking process and provide users with more accurate results. | cs.CL | PAQA: Toward ProActive Open-Retrieval Question Answering |
2024-02-26T14:30:37Z | http://arxiv.org/pdf/2402.16602v2 | http://arxiv.org/abs/2402.16602v2 | Yuyang Ding, Juntao Li, Pinzheng Wang, Zecheng Tang, Bowen Yan, Min Zhang | 2024-06-19T03:16:58Z | cs.CL | Large Language Models (LLMs) have demonstrated impressive capabilities for
generalizing in unseen tasks. In the Named Entity Recognition (NER) task,
recent advancements have seen the remarkable improvement of LLMs in a broad
range of entity domains via instruction tuning, by adopting entity-centric
schema. In this work, we explore the potential enhancement of the existing
methods by incorporating negative instances into training. Our experiments
reveal that negative instances contribute to remarkable improvements by (1)
introducing contextual information, and (2) clearly delineating label
boundaries. Furthermore, we introduce an efficient longest common subsequence
(LCS) matching algorithm, which is tailored to transform unstructured
predictions into structured entities. By integrating these components, we
present GNER, a Generative NER system that shows improved zero-shot performance
across unseen entity domains. Our comprehensive evaluation illustrates our
system's superiority, surpassing state-of-the-art (SoTA) methods by 9 $F_1$
score in zero-shot evaluation. | cs.CL | Rethinking Negative Instances for Generative Named Entity Recognition |
2024-02-26T14:27:06Z | http://arxiv.org/pdf/2402.16596v1 | http://arxiv.org/abs/2402.16596v1 | Marko Pranjić, Kaja Dobrovoljc, Senja Pollak, Matej Martinc | 2024-02-26T14:27:06Z | cs.CL, I.2.7 | In this paper, we focus on the detection of semantic changes in Slovene, a
less resourced Slavic language with two million speakers. Detecting and
tracking semantic changes provides insights into the evolution of the language
caused by changes in society and culture. Recently, several systems have been
proposed to aid in this study, but all depend on manually annotated gold
standard datasets for evaluation. In this paper, we present the first Slovene
dataset for evaluating semantic change detection systems, which contains
aggregated semantic change scores for 104 target words obtained from more than
3000 manually annotated sentence pairs. We evaluate several existing semantic
change detection methods on this dataset and also propose a novel approach
based on optimal transport that improves on the existing state-of-the-art
systems with an error reduction rate of 22.8%. | cs.CL | Semantic change detection for Slovene language: a novel dataset and an approach based on optimal transport |
2024-02-26T14:01:34Z | http://arxiv.org/pdf/2402.16578v1 | http://arxiv.org/abs/2402.16578v1 | Massieh Kordi Boroujeny, Ya Jiang, Kai Zeng, Brian Mark | 2024-02-26T14:01:34Z | cs.CL, cs.LG | Methods for watermarking large language models have been proposed that
distinguish AI-generated text from human-generated text by slightly altering
the model output distribution, but they also distort the quality of the text,
exposing the watermark to adversarial detection. More recently, distortion-free
watermarking methods were proposed that require a secret key to detect the
watermark. The prior methods generally embed zero-bit watermarks that do not
provide additional information beyond tagging a text as being AI-generated. We
extend an existing zero-bit distortion-free watermarking method by embedding
multiple bits of meta-information as part of the watermark. We also develop a
computationally efficient decoder that extracts the embedded information from
the watermark with low bit error rate. | cs.CL | Multi-Bit Distortion-Free Watermarking for Large Language Models |
2024-02-26T13:47:09Z | http://arxiv.org/pdf/2402.16568v1 | http://arxiv.org/abs/2402.16568v1 | Yifu Gao, Linbo Qiao, Zhigang Kan, Zhihua Wen, Yongquan He, Dongsheng Li | 2024-02-26T13:47:09Z | cs.CL | Temporal knowledge graph question answering (TKGQA) poses a significant
challenge task, due to the temporal constraints hidden in questions and the
answers sought from dynamic structured knowledge. Although large language
models (LLMs) have made considerable progress in their reasoning ability over
structured data, their application to the TKGQA task is a relatively unexplored
area. This paper first proposes a novel generative temporal knowledge graph
question answering framework, GenTKGQA, which guides LLMs to answer temporal
questions through two phases: Subgraph Retrieval and Answer Generation. First,
we exploit LLM's intrinsic knowledge to mine temporal constraints and
structural links in the questions without extra training, thus narrowing down
the subgraph search space in both temporal and structural dimensions. Next, we
design virtual knowledge indicators to fuse the graph neural network signals of
the subgraph and the text representations of the LLM in a non-shallow way,
which helps the open-source LLM deeply understand the temporal order and
structural dependencies among the retrieved facts through instruction tuning.
Experimental results demonstrate that our model outperforms state-of-the-art
baselines, even achieving 100\% on the metrics for the simple question type. | cs.CL | Two-stage Generative Question Answering on Temporal Knowledge Graph Using Large Language Models |
2024-02-26T13:46:51Z | http://arxiv.org/pdf/2402.16567v2 | http://arxiv.org/abs/2402.16567v2 | Yuanyuan Liang, Keren Tan, Tingyu Xie, Wenbiao Tao, Siyuan Wang, Yunshi Lan, Weining Qian | 2024-02-28T07:24:19Z | cs.CL, cs.AI, cs.DB | Graph Databases (Graph DB) are widely applied in various fields, including
finance, social networks, and medicine. However, translating Natural Language
(NL) into the Graph Query Language (GQL), commonly known as NL2GQL, proves to
be challenging due to its inherent complexity and specialized nature. Some
approaches have sought to utilize Large Language Models (LLMs) to address
analogous tasks like text2SQL. Nevertheless, when it comes to NL2GQL taskson a
particular domain, the absence of domain-specific NL-GQL data pairs makes it
difficult to establish alignment between LLMs and the graph DB. To address this
challenge, we propose a well-defined pipeline. Specifically, we utilize ChatGPT
to create NL-GQL data pairs based on the given graph DB with self-instruct.
Then, we use the created data to fine-tune LLMs, thereby achieving alignment
between LLMs and the graph DB. Additionally, during inference, we propose a
method that extracts relevant schema to the queried NL as the input context to
guide LLMs for generating accurate GQLs.We evaluate our method on two
constructed datasets deriving from graph DBs in finance domain and medicine
domain, namely FinGQL and MediGQL. Experimental results demonstrate that our
method significantly outperforms a set of baseline methods, with improvements
of 5.90 and 6.36 absolute points on EM, and 6.00 and 7.09 absolute points on
EX, respectively. | cs.CL | Aligning Large Language Models to a Domain-specific Graph Database |
2024-02-26T12:56:17Z | http://arxiv.org/pdf/2403.00820v1 | http://arxiv.org/abs/2403.00820v1 | Tristan Kenneweg, Philip Kenneweg, Barbara Hammer | 2024-02-26T12:56:17Z | cs.IR, cs.CL | Retrieval Augmented Generation (RAG) systems have seen huge popularity in
augmenting Large-Language Model (LLM) outputs with domain specific and time
sensitive data. Very recently a shift is happening from simple RAG setups that
query a vector database for additional information with every user input to
more sophisticated forms of RAG. However, different concrete approaches compete
on mostly anecdotal evidence at the moment. In this paper we present a rigorous
dataset creation and evaluation workflow to quantitatively compare different
RAG strategies. We use a dataset created this way for the development and
evaluation of a boolean agent RAG setup: A system in which a LLM can decide
whether to query a vector database or not, thus saving tokens on questions that
can be answered with internal knowledge. We publish our code and generated
dataset online. | cs.IR | Retrieval Augmented Generation Systems: Automatic Dataset Creation, Evaluation and Boolean Agent Setup |
2024-02-26T12:55:51Z | http://arxiv.org/pdf/2402.16539v1 | http://arxiv.org/abs/2402.16539v1 | Naicheng Guo, Hongwei Cheng, Qianqiao Liang, Linxun Chen, Bing Han | 2024-02-26T12:55:51Z | cs.IR, cs.CL, cs.LG | With the rapid development of Large Language Models (LLMs), various
explorations have arisen to utilize LLMs capability of context understanding on
recommender systems. While pioneering strategies have primarily transformed
traditional recommendation tasks into challenges of natural language
generation, there has been a relative scarcity of exploration in the domain of
session-based recommendation (SBR) due to its specificity. SBR has been
primarily dominated by Graph Neural Networks, which have achieved many
successful outcomes due to their ability to capture both the implicit and
explicit relationships between adjacent behaviors. The structural nature of
graphs contrasts with the essence of natural language, posing a significant
adaptation gap for LLMs. In this paper, we introduce large language models with
graphical Session-Based recommendation, named LLMGR, an effective framework
that bridges the aforementioned gap by harmoniously integrating LLMs with Graph
Neural Networks (GNNs) for SBR tasks. This integration seeks to leverage the
complementary strengths of LLMs in natural language understanding and GNNs in
relational data processing, leading to a more powerful session-based
recommender system that can understand and recommend items within a session.
Moreover, to endow the LLM with the capability to empower SBR tasks, we design
a series of prompts for both auxiliary and major instruction tuning tasks.
These prompts are crafted to assist the LLM in understanding graph-structured
data and align textual information with nodes, effectively translating nuanced
user interactions into a format that can be understood and utilized by LLM
architectures. Extensive experiments on three real-world datasets demonstrate
that LLMGR outperforms several competitive baselines, indicating its
effectiveness in enhancing SBR tasks and its potential as a research direction
for future exploration. | cs.IR | Integrating Large Language Models with Graphical Session-Based Recommendation |
2024-02-26T11:52:55Z | http://arxiv.org/pdf/2402.16515v1 | http://arxiv.org/abs/2402.16515v1 | Yiping Song, Juhua Zhang, Zhiliang Tian, Yuxin Yang, Minlie Huang, Dongsheng Li | 2024-02-26T11:52:55Z | cs.CL, cs.CR | As sufficient data are not always publically accessible for model training,
researchers exploit limited data with advanced learning algorithms or expand
the dataset via data augmentation (DA). Conducting DA in private domain
requires private protection approaches (i.e. anonymization and perturbation),
but those methods cannot provide protection guarantees. Differential privacy
(DP) learning methods theoretically bound the protection but are not skilled at
generating pseudo text samples with large models. In this paper, we transfer
DP-based pseudo sample generation task to DP-based generated samples
discrimination task, where we propose a DP-based DA method with a LLM and a
DP-based discriminator for text classification on private domains. We construct
a knowledge distillation model as the DP-based discriminator: teacher models,
accessing private data, teaches students how to select private samples with
calibrated noise to achieve DP. To constrain the distribution of DA's
generation, we propose a DP-based tutor that models the noised private
distribution and controls samples' generation with a low privacy cost. We
theoretically analyze our model's privacy protection and empirically verify our
model. | cs.CL | LLM-based Privacy Data Augmentation Guided by Knowledge Distillation with a Distribution Tutor for Medical Text Classification |
2024-02-26T11:42:29Z | http://arxiv.org/pdf/2402.16508v2 | http://arxiv.org/abs/2402.16508v2 | Fan Jiang, Tom Drummond, Trevor Cohn | 2024-06-16T08:18:25Z | cs.CL, cs.IR | Cross-lingual open domain question answering (CLQA) is a complex problem,
comprising cross-lingual retrieval from a multilingual knowledge base, followed
by answer generation in the query language. Both steps are usually tackled by
separate models, requiring substantial annotated datasets, and typically
auxiliary resources, like machine translation systems to bridge between
languages. In this paper, we show that CLQA can be addressed using a single
encoder-decoder model. To effectively train this model, we propose a
self-supervised method based on exploiting the cross-lingual link structure
within Wikipedia. We demonstrate how linked Wikipedia pages can be used to
synthesise supervisory signals for cross-lingual retrieval, through a form of
cloze query, and generate more natural questions to supervise answer
generation. Together, we show our approach, \texttt{CLASS}, outperforms
comparable methods on both supervised and zero-shot language adaptation
settings, including those using machine translation. | cs.CL | Pre-training Cross-lingual Open Domain Question Answering with Large-scale Synthetic Supervision |
2024-02-26T11:31:48Z | http://arxiv.org/pdf/2402.16499v1 | http://arxiv.org/abs/2402.16499v1 | Junzhe Chen, Xuming Hu, Shuodi Liu, Shiyu Huang, Wei-Wei Tu, Zhaofeng He, Lijie Wen | 2024-02-26T11:31:48Z | cs.CL | Recent advancements in large language models (LLMs) have revealed their
potential for achieving autonomous agents possessing human-level intelligence.
However, existing benchmarks for evaluating LLM Agents either use static
datasets, potentially leading to data leakage or focus only on single-agent
scenarios, overlooking the complexities of multi-agent interactions. There is a
lack of a benchmark that evaluates the diverse capabilities of LLM agents in
multi-agent, dynamic environments. To this end, we introduce LLMArena, a novel
and easily extensible framework for evaluating the diverse capabilities of LLM
in multi-agent dynamic environments. LLMArena encompasses seven distinct gaming
environments, employing Trueskill scoring to assess crucial abilities in LLM
agents, including spatial reasoning, strategic planning, numerical reasoning,
risk assessment, communication, opponent modeling, and team collaboration. We
conduct an extensive experiment and human evaluation among different sizes and
types of LLMs, showing that LLMs still have a significant journey ahead in
their development towards becoming fully autonomous agents, especially in
opponent modeling and team collaboration. We hope LLMArena could guide future
research towards enhancing these capabilities in LLMs, ultimately leading to
more sophisticated and practical applications in dynamic, multi-agent settings.
The code and data will be available. | cs.CL | LLMArena: Assessing Capabilities of Large Language Models in Dynamic Multi-Agent Environments |
2024-02-26T11:01:54Z | http://arxiv.org/pdf/2402.16482v1 | http://arxiv.org/abs/2402.16482v1 | Han Liu, Liantang Li | 2024-02-26T11:01:54Z | cs.AI, cs.CE, cs.CL | Language model intelligence is revolutionizing the way we program materials
simulations. However, the diversity of simulation scenarios renders it
challenging to precisely transform human language into a tailored simulator.
Here, using three functionalized types of language model, we propose a
language-to-simulation (Lang2Sim) framework that enables interactive navigation
on languaging a simulation engine, by taking a scenario instance of water
sorption in porous matrices. Unlike line-by-line coding of a target simulator,
the language models interpret each simulator as an assembly of invariant tool
function and its variant input-output pair. Lang2Sim enables the precise
transform of textual description by functionalizing and sequentializing the
language models of, respectively, rationalizing the tool categorization,
customizing its input-output combinations, and distilling the simulator input
into executable format. Importantly, depending on its functionalized type, each
language model features a distinct processing of chat history to best balance
its memory limit and information completeness, thus leveraging the model
intelligence to unstructured nature of human request. Overall, this work
establishes language model as an intelligent platform to unlock the era of
languaging a simulation engine. | cs.AI | On Languaging a Simulation Engine |
2024-02-26T10:33:36Z | http://arxiv.org/pdf/2402.16472v2 | http://arxiv.org/abs/2402.16472v2 | Vipul Raheja, Dimitris Alikaniotis, Vivek Kulkarni, Bashar Alhafni, Dhruv Kumar | 2024-04-17T16:59:30Z | cs.CL, cs.AI, I.2.7 | We introduce mEdIT, a multi-lingual extension to CoEdIT -- the recent
state-of-the-art text editing models for writing assistance. mEdIT models are
trained by fine-tuning multi-lingual large, pre-trained language models (LLMs)
via instruction tuning. They are designed to take instructions from the user
specifying the attributes of the desired text in the form of natural language
instructions, such as Grammatik korrigieren (German) or Parafrasee la oraci\'on
(Spanish). We build mEdIT by curating data from multiple publicly available
human-annotated text editing datasets for three text editing tasks (Grammatical
Error Correction (GEC), Text Simplification, and Paraphrasing) across diverse
languages belonging to six different language families. We detail the design
and training of mEdIT models and demonstrate their strong performance on many
multi-lingual text editing benchmarks against other multilingual LLMs. We also
find that mEdIT generalizes effectively to new languages over multilingual
baselines. We publicly release our data, code, and trained models at
https://github.com/vipulraheja/medit. | cs.CL | mEdIT: Multilingual Text Editing via Instruction Tuning |
2024-02-26T10:31:45Z | http://arxiv.org/pdf/2402.16470v1 | http://arxiv.org/abs/2402.16470v1 | Khai Jiet Liong, Hongqiu Wu, Hai Zhao | 2024-02-26T10:31:45Z | cs.CL | Pre-trained language models (PLMs) are shown to be vulnerable to minor word
changes, which poses a big threat to real-world systems. While previous studies
directly focus on manipulating word inputs, they are limited by their means of
generating adversarial samples, lacking generalization to versatile real-world
attack. This paper studies the basic structure of transformer-based PLMs, the
self-attention (SA) mechanism. (1) We propose a powerful perturbation technique
\textit{HackAttend}, which perturbs the attention scores within the SA matrices
via meticulously crafted attention masks. We show that state-of-the-art PLMs
fall into heavy vulnerability that minor attention perturbations $(1\%)$ can
produce a very high attack success rate $(98\%)$. Our paper expands the
conventional text attack of word perturbations to more general structural
perturbations. (2) We introduce \textit{S-Attend}, a novel smoothing technique
that effectively makes SA robust via structural perturbations. We empirically
demonstrate that this simple yet effective technique achieves robust
performance on par with adversarial training when facing various text
attackers. Code is publicly available at \url{github.com/liongkj/HackAttend}. | cs.CL | Unveiling Vulnerability of Self-Attention |
2024-02-26T10:03:33Z | http://arxiv.org/pdf/2402.16459v3 | http://arxiv.org/abs/2402.16459v3 | Yihan Wang, Zhouxing Shi, Andrew Bai, Cho-Jui Hsieh | 2024-06-06T19:21:44Z | cs.CL, cs.AI | Although many large language models (LLMs) have been trained to refuse
harmful requests, they are still vulnerable to jailbreaking attacks which
rewrite the original prompt to conceal its harmful intent. In this paper, we
propose a new method for defending LLMs against jailbreaking attacks by
``backtranslation''. Specifically, given an initial response generated by the
target LLM from an input prompt, our backtranslation prompts a language model
to infer an input prompt that can lead to the response. The inferred prompt is
called the backtranslated prompt which tends to reveal the actual intent of the
original prompt, since it is generated based on the LLM's response and not
directly manipulated by the attacker. We then run the target LLM again on the
backtranslated prompt, and we refuse the original prompt if the model refuses
the backtranslated prompt. We explain that the proposed defense provides
several benefits on its effectiveness and efficiency. We empirically
demonstrate that our defense significantly outperforms the baselines, in the
cases that are hard for the baselines, and our defense also has little impact
on the generation quality for benign input prompts. Our implementation is based
on our library for LLM jailbreaking defense algorithms at
\url{https://github.com/YihanWang617/llm-jailbreaking-defense}, and the code
for reproducing our experiments is available at
\url{https://github.com/YihanWang617/LLM-Jailbreaking-Defense-Backtranslation}. | cs.CL | Defending LLMs against Jailbreaking Attacks via Backtranslation |
2024-02-26T10:02:29Z | http://arxiv.org/pdf/2402.16458v2 | http://arxiv.org/abs/2402.16458v2 | Peiling Yi, Arkaitz Zubiaga | 2024-02-27T12:52:24Z | cs.CL | Swear words are a common proxy to collect datasets with cyberbullying
incidents. Our focus is on measuring and mitigating biases derived from
spurious associations between swear words and incidents occurring as a result
of such data collection strategies. After demonstrating and quantifying these
biases, we introduce ID-XCB, the first data-independent debiasing technique
that combines adversarial training, bias constraints and debias fine-tuning
approach aimed at alleviating model attention to bias-inducing words without
impacting overall model performance. We explore ID-XCB on two popular
session-based cyberbullying datasets along with comprehensive ablation and
generalisation studies. We show that ID-XCB learns robust cyberbullying
detection capabilities while mitigating biases, outperforming state-of-the-art
debiasing methods in both performance and bias mitigation. Our quantitative and
qualitative analyses demonstrate its generalisability to unseen data. | cs.CL | ID-XCB: Data-independent Debiasing for Fair and Accurate Transformer-based Cyberbullying Detection |
2024-02-26T09:59:04Z | http://arxiv.org/pdf/2402.16457v2 | http://arxiv.org/abs/2402.16457v2 | Zihan Zhang, Meng Fang, Ling Chen | 2024-06-05T05:23:21Z | cs.CL | Adaptive retrieval-augmented generation (ARAG) aims to dynamically determine
the necessity of retrieval for queries instead of retrieving indiscriminately
to enhance the efficiency and relevance of the sourced information. However,
previous works largely overlook the evaluation of ARAG approaches, leading to
their effectiveness being understudied. This work presents a benchmark,
RetrievalQA, comprising 1,271 short-form questions covering new world and
long-tail knowledge. The knowledge necessary to answer the questions is absent
from LLMs; therefore, external information must be retrieved to answer
correctly. This makes RetrievalQA a suitable testbed to evaluate existing ARAG
methods. We observe that calibration-based methods heavily rely on threshold
tuning, while vanilla prompting is inadequate for guiding LLMs to make reliable
retrieval decisions. Based on our findings, we propose Time-Aware Adaptive
Retrieval (TA-ARE), a simple yet effective method that helps LLMs assess the
necessity of retrieval without calibration or additional training. The dataset
and code will be available at https://github.com/hyintell/RetrievalQA | cs.CL | RetrievalQA: Assessing Adaptive Retrieval-Augmented Generation for Short-form Open-Domain Question Answering |
2024-02-26T09:43:02Z | http://arxiv.org/pdf/2402.16444v1 | http://arxiv.org/abs/2402.16444v1 | Zhexin Zhang, Yida Lu, Jingyuan Ma, Di Zhang, Rui Li, Pei Ke, Hao Sun, Lei Sha, Zhifang Sui, Hongning Wang, Minlie Huang | 2024-02-26T09:43:02Z | cs.CL | The safety of Large Language Models (LLMs) has gained increasing attention in
recent years, but there still lacks a comprehensive approach for detecting
safety issues within LLMs' responses in an aligned, customizable and
explainable manner. In this paper, we propose ShieldLM, an LLM-based safety
detector, which aligns with general human safety standards, supports
customizable detection rules, and provides explanations for its decisions. To
train ShieldLM, we compile a large bilingual dataset comprising 14,387
query-response pairs, annotating the safety of responses based on various
safety standards. Through extensive experiments, we demonstrate that ShieldLM
surpasses strong baselines across four test sets, showcasing remarkable
customizability and explainability. Besides performing well on standard
detection datasets, ShieldLM has also been shown to be effective in real-world
situations as a safety evaluator for advanced LLMs. We release ShieldLM at
\url{https://github.com/thu-coai/ShieldLM} to support accurate and explainable
safety detection under various safety standards, contributing to the ongoing
efforts to enhance the safety of LLMs. | cs.CL | ShieldLM: Empowering LLMs as Aligned, Customizable and Explainable Safety Detectors |
2024-02-26T09:36:05Z | http://arxiv.org/pdf/2402.16438v2 | http://arxiv.org/abs/2402.16438v2 | Tianyi Tang, Wenyang Luo, Haoyang Huang, Dongdong Zhang, Xiaolei Wang, Xin Zhao, Furu Wei, Ji-Rong Wen | 2024-06-06T08:14:46Z | cs.CL | Large language models (LLMs) demonstrate remarkable multilingual capabilities
without being pre-trained on specially curated multilingual parallel corpora.
It remains a challenging problem to explain the underlying mechanisms by which
LLMs process multilingual texts. In this paper, we delve into the composition
of Transformer architectures in LLMs to pinpoint language-specific regions.
Specially, we propose a novel detection method, language activation probability
entropy (LAPE), to identify language-specific neurons within LLMs. Based on
LAPE, we conduct comprehensive experiments on several representative LLMs, such
as LLaMA-2, BLOOM, and Mistral. Our findings indicate that LLMs' proficiency in
processing a particular language is predominantly due to a small subset of
neurons, primarily situated in the models' top and bottom layers. Furthermore,
we showcase the feasibility to "steer" the output language of LLMs by
selectively activating or deactivating language-specific neurons. Our research
provides important evidence to the understanding and exploration of the
multilingual capabilities of LLMs. | cs.CL | Language-Specific Neurons: The Key to Multilingual Capabilities in Large Language Models |
2024-02-26T09:30:55Z | http://arxiv.org/pdf/2402.16431v1 | http://arxiv.org/abs/2402.16431v1 | Yuansen Zhang, Xiao Wang, Zhiheng Xi, Han Xia, Tao Gui, Qi Zhang, Xuanjing Huang | 2024-02-26T09:30:55Z | cs.CL | Large Language Models (LLMs) have showcased remarkable capabilities in
following human instructions. However, recent studies have raised concerns
about the robustness of LLMs when prompted with instructions combining textual
adversarial samples. In this paper, drawing inspiration from recent works that
LLMs are sensitive to the design of the instructions, we utilize instructions
in code style, which are more structural and less ambiguous, to replace
typically natural language instructions. Through this conversion, we provide
LLMs with more precise instructions and strengthen the robustness of LLMs.
Moreover, under few-shot scenarios, we propose a novel method to compose
in-context demonstrations using both clean and adversarial samples
(\textit{adversarial context method}) to further boost the robustness of the
LLMs. Experiments on eight robustness datasets show that our method
consistently outperforms prompting LLMs with natural language instructions. For
example, with gpt-3.5-turbo, our method achieves an improvement of 5.68\% in
test set accuracy and a reduction of 5.66 points in Attack Success Rate (ASR). | cs.CL | RoCoIns: Enhancing Robustness of Large Language Models through Code-Style Instructions |
2024-02-26T09:21:59Z | http://arxiv.org/pdf/2403.00818v2 | http://arxiv.org/abs/2403.00818v2 | Wei He, Kai Han, Yehui Tang, Chengcheng Wang, Yujie Yang, Tianyu Guo, Yunhe Wang | 2024-03-05T14:31:03Z | cs.CL, cs.LG | Large language models (LLMs) face a daunting challenge due to the excessive
computational and memory requirements of the commonly used Transformer
architecture. While state space model (SSM) is a new type of foundational
network architecture offering lower computational complexity, their performance
has yet to fully rival that of Transformers. This paper introduces DenseSSM, a
novel approach to enhance the flow of hidden information between layers in
SSMs. By selectively integrating shallowlayer hidden states into deeper layers,
DenseSSM retains fine-grained information crucial for the final output. Dense
connections enhanced DenseSSM still maintains the training parallelizability
and inference efficiency. The proposed method can be widely applicable to
various SSM types like RetNet and Mamba. With similar model size, DenseSSM
achieves significant improvements, exemplified by DenseRetNet outperforming the
original RetNet with up to 5% accuracy improvement on public benchmarks. code
is avalaible at https://github.com/WailordHe/DenseSSM | cs.CL | DenseMamba: State Space Models with Dense Hidden Connection for Efficient Large Language Models |
2024-02-26T09:19:46Z | http://arxiv.org/pdf/2402.16420v2 | http://arxiv.org/abs/2402.16420v2 | Lev Kharlashkin, Melany Macias, Leo Huovinen, Mika Hämäläinen | 2024-04-23T12:49:57Z | cs.CL | We present our work on predicting United Nations sustainable development
goals (SDG) for university courses. We use an LLM named PaLM 2 to generate
training data given a noisy human-authored course description input as input.
We use this data to train several different smaller language models to predict
SDGs for university courses. This work contributes to better university level
adaptation of SDGs. The best performing model in our experiments was BART with
an F1-score of 0.786. | cs.CL | Predicting Sustainable Development Goals Using Course Descriptions -- from LLMs to Conventional Foundation Models |
2024-02-26T09:10:34Z | http://arxiv.org/pdf/2403.05574v2 | http://arxiv.org/abs/2403.05574v2 | Mengxi Xiao, Qianqian Xie, Ziyan Kuang, Zhicheng Liu, Kailai Yang, Min Peng, Weiguang Han, Jimin Huang | 2024-03-22T07:34:38Z | cs.HC, cs.AI, cs.CL, J.4 | Large Language Models (LLMs) can play a vital role in psychotherapy by
adeptly handling the crucial task of cognitive reframing and overcoming
challenges such as shame, distrust, therapist skill variability, and resource
scarcity. Previous LLMs in cognitive reframing mainly converted negative
emotions to positive ones, but these approaches have limited efficacy, often
not promoting clients' self-discovery of alternative perspectives. In this
paper, we unveil the Helping and Empowering through Adaptive Language in Mental
Enhancement (HealMe) model. This novel cognitive reframing therapy method
effectively addresses deep-rooted negative thoughts and fosters rational,
balanced perspectives. Diverging from traditional LLM methods, HealMe employs
empathetic dialogue based on psychotherapeutic frameworks. It systematically
guides clients through distinguishing circumstances from feelings,
brainstorming alternative viewpoints, and developing empathetic, actionable
suggestions. Moreover, we adopt the first comprehensive and expertly crafted
psychological evaluation metrics, specifically designed to rigorously assess
the performance of cognitive reframing, in both AI-simulated dialogues and
real-world therapeutic conversations. Experimental results show that our model
outperforms others in terms of empathy, guidance, and logical coherence,
demonstrating its effectiveness and potential positive impact on psychotherapy. | cs.HC | HealMe: Harnessing Cognitive Reframing in Large Language Models for Psychotherapy |
2024-02-26T08:59:05Z | http://arxiv.org/pdf/2402.16406v1 | http://arxiv.org/abs/2402.16406v1 | Nigel Markey, Ilyass El-Mansouri, Gaetan Rensonnet, Casper van Langen, Christoph Meier | 2024-02-26T08:59:05Z | cs.CL | Clinical trials require numerous documents to be written -- protocols,
consent forms, clinical study reports and others. Large language models (LLMs)
offer the potential to rapidly generate first versions of these documents,
however there are concerns about the quality of their output Here we report an
evaluation of LLMs in generating parts of one such document, clinical trial
protocols. We find that an offthe-shelf LLM delivers reasonable results,
especially when assessing content relevance and the correct use of terminology.
However, deficiencies remain: specifically clinical thinking and logic, and
appropriate use of references. To improve performance, we used
retrieval-augmented generation (RAG) to prompt an LLM with accurate up-to-date
information. As a result of using RAG, the writing quality of the LLM improves
substantially, which has implications for the practical useability of LLMs in
clinical trial-related writing. | cs.CL | From RAGs to riches: Using large language models to write documents for clinical trials |
2024-02-26T08:27:50Z | http://arxiv.org/pdf/2402.16389v1 | http://arxiv.org/abs/2402.16389v1 | Shiwen Ni, Minghuan Tan, Yuelin Bai, Fuqiang Niu, Min Yang, Bowen Zhang, Ruifeng Xu, Xiaojun Chen, Chengming Li, Xiping Hu, Ye Li, Jianping Fan | 2024-02-26T08:27:50Z | cs.CL, cs.AI | Large language models (LLMs) have demonstrated impressive performance in
various natural language processing (NLP) tasks. However, there is limited
understanding of how well LLMs perform in specific domains (e.g, the
intellectual property (IP) domain). In this paper, we contribute a new
benchmark, the first Multilingual-oriented quiZ on Intellectual Property
(MoZIP), for the evaluation of LLMs in the IP domain. The MoZIP benchmark
includes three challenging tasks: IP multiple-choice quiz (IPQuiz), IP question
answering (IPQA), and patent matching (PatentMatch). In addition, we also
develop a new IP-oriented multilingual large language model (called MoZi),
which is a BLOOMZ-based model that has been supervised fine-tuned with
multilingual IP-related text data. We evaluate our proposed MoZi model and four
well-known LLMs (i.e., BLOOMZ, BELLE, ChatGLM and ChatGPT) on the MoZIP
benchmark. Experimental results demonstrate that MoZi outperforms BLOOMZ, BELLE
and ChatGLM by a noticeable margin, while it had lower scores compared with
ChatGPT. Notably, the performance of current LLMs on the MoZIP benchmark has
much room for improvement, and even the most powerful ChatGPT does not reach
the passing level. Our source code, data, and models are available at
\url{https://github.com/AI-for-Science/MoZi}. | cs.CL | MoZIP: A Multilingual Benchmark to Evaluate Large Language Models in Intellectual Property |
2024-02-26T08:08:03Z | http://arxiv.org/pdf/2402.16382v1 | http://arxiv.org/abs/2402.16382v1 | Domenic Rosati, Jan Wehner, Kai Williams, Łukasz Bartoszcze, Jan Batzner, Hassan Sajjad, Frank Rudzicz | 2024-02-26T08:08:03Z | cs.CL | Approaches to aligning large language models (LLMs) with human values has
focused on correcting misalignment that emerges from pretraining. However, this
focus overlooks another source of misalignment: bad actors might purposely
fine-tune LLMs to achieve harmful goals. In this paper, we present an emerging
threat model that has arisen from alignment circumvention and fine-tuning
attacks. However, lacking in previous works is a clear presentation of the
conditions for effective defence. We propose a set of conditions for effective
defence against harmful fine-tuning in LLMs called "Immunization conditions,"
which help us understand how we would construct and measure future defences.
Using this formal framework for defence, we offer a synthesis of different
research directions that might be persued to prevent harmful fine-tuning
attacks and provide a demonstration of how to use these conditions
experimentally showing early results of using an adversarial loss to immunize
LLama2-7b-chat. | cs.CL | Immunization against harmful fine-tuning attacks |
2024-02-26T07:58:33Z | http://arxiv.org/pdf/2402.16380v1 | http://arxiv.org/abs/2402.16380v1 | Ahmet Gunduz, Kamer Ali Yuksel, Kareem Darwish, Golara Javadi, Fabio Minazzi, Nicola Sobieski, Sebastien Bratieres | 2024-02-26T07:58:33Z | eess.AS, cs.AI, cs.CL, cs.LG | Data availability is crucial for advancing artificial intelligence
applications, including voice-based technologies. As content creation,
particularly in social media, experiences increasing demand, translation and
text-to-speech (TTS) technologies have become essential tools. Notably, the
performance of these TTS technologies is highly dependent on the quality of the
training data, emphasizing the mutual dependence of data availability and
technological progress. This paper introduces an end-to-end tool to generate
high-quality datasets for text-to-speech (TTS) models to address this critical
need for high-quality data. The contributions of this work are manifold and
include: the integration of language-specific phoneme distribution into sample
selection, automation of the recording process, automated and human-in-the-loop
quality assurance of recordings, and processing of recordings to meet specified
formats. The proposed application aims to streamline the dataset creation
process for TTS models through these features, thereby facilitating
advancements in voice-based technologies. | eess.AS | An Automated End-to-End Open-Source Software for High-Quality Text-to-Speech Dataset Generation |
2024-02-26T07:58:12Z | http://arxiv.org/pdf/2402.16379v3 | http://arxiv.org/abs/2402.16379v3 | Zhaopeng Feng, Yan Zhang, Hao Li, Bei Wu, Jiayu Liao, Wenqiang Liu, Jun Lang, Yang Feng, Jian Wu, Zuozhu Liu | 2024-06-21T07:35:53Z | cs.CL, cs.AI | Large Language Models (LLMs) have achieved impressive results in Machine
Translation (MT). However, careful evaluations by human reveal that the
translations produced by LLMs still contain multiple errors. Importantly,
feeding back such error information into the LLMs can lead to self-refinement
and result in improved translation performance. Motivated by these insights, we
introduce a systematic LLM-based self-refinement translation framework, named
\textbf{TEaR}, which stands for \textbf{T}ranslate, \textbf{E}stimate,
\textbf{a}nd \textbf{R}efine, marking a significant step forward in this
direction. Our findings demonstrate that 1) our self-refinement framework
successfully assists LLMs in improving their translation quality across a wide
range of languages, whether it's from high-resource languages to low-resource
ones or whether it's English-centric or centered around other languages; 2)
TEaR exhibits superior systematicity and interpretability; 3) different
estimation strategies yield varied impacts, directly affecting the
effectiveness of the final corrections. Additionally, traditional neural
translation models and evaluation models operate separately, often focusing on
singular tasks due to their limited capabilities, while general-purpose LLMs
possess the capability to undertake both tasks simultaneously. We further
conduct cross-model correction experiments to investigate the potential
relationship between the translation and evaluation capabilities of
general-purpose LLMs. Our code and data are available at
https://github.com/fzp0424/self_correct_mt | cs.CL | TEaR: Improving LLM-based Machine Translation with Systematic Self-Refinement |
2024-02-26T07:44:56Z | http://arxiv.org/pdf/2402.16367v2 | http://arxiv.org/abs/2402.16367v2 | Weize Liu, Yinlong Xu, Hongxia Xu, Jintai Chen, Xuming Hu, Jian Wu | 2024-06-17T17:57:32Z | cs.CL | Recently, large language models (LLMs) have achieved tremendous breakthroughs
in the field of NLP, but still lack understanding of their internal activities
when processing different languages. We designed a method to convert dense LLMs
into fine-grained MoE architectures, and then visually studied the multilingual
activation patterns of LLMs through expert activation frequency heatmaps.
Through comprehensive experiments on different model families, different model
sizes, and different variants, we analyzed the distribution of high-frequency
activated experts, multilingual shared experts, whether the activation patterns
of different languages are related to language families, and the impact of
instruction tuning on activation patterns. We further explored leveraging the
discovered differences in expert activation frequencies to guide unstructured
pruning in two different ways. Experimental results demonstrated that our
method significantly outperformed random expert pruning and even exceeded the
performance of the original unpruned models in some languages. Additionally, we
found that configuring different pruning rates for different layers based on
activation level differences could achieve better results. Our findings reveal
the multilingual processing mechanisms within LLMs and utilize these insights
to offer new perspectives for applications such as model pruning. | cs.CL | Unraveling Babel: Exploring Multilingual Activation Patterns of LLMs and Their Applications |
2024-02-26T07:33:28Z | http://arxiv.org/pdf/2402.16364v1 | http://arxiv.org/abs/2402.16364v1 | Tzuf Paz-Argaman, Sayali Kulkarni, John Palowitch, Jason Baldridge, Reut Tsarfaty | 2024-02-26T07:33:28Z | cs.CL, cs.LG, cs.MM | When communicating routes in natural language, the concept of {\em acquired
spatial knowledge} is crucial for geographic information retrieval (GIR) and in
spatial cognitive research. However, NLP navigation studies often overlook the
impact of such acquired knowledge on textual descriptions. Current navigation
studies concentrate on egocentric local descriptions (e.g., `it will be on your
right') that require reasoning over the agent's local perception. These
instructions are typically given as a sequence of steps, with each action-step
explicitly mentioning and being followed by a landmark that the agent can use
to verify they are on the right path (e.g., `turn right and then you will
see...'). In contrast, descriptions based on knowledge acquired through a map
provide a complete view of the environment and capture its overall structure.
These instructions (e.g., `it is south of Central Park and a block north of a
police station') are typically non-sequential, contain allocentric relations,
with multiple spatial relations and implicit actions, without any explicit
verification. This paper introduces the Rendezvous (RVS) task and dataset,
which includes 10,404 examples of English geospatial instructions for reaching
a target location using map-knowledge. Our analysis reveals that RVS exhibits a
richer use of spatial allocentric relations, and requires resolving more
spatial relations simultaneously compared to previous text-based navigation
benchmarks. | cs.CL | Where Do We Go from Here? Multi-scale Allocentric Relational Inference from Natural Spatial Descriptions |
2024-02-26T07:33:05Z | http://arxiv.org/pdf/2402.16363v6 | http://arxiv.org/abs/2402.16363v6 | Zhihang Yuan, Yuzhang Shang, Yang Zhou, Zhen Dong, Zhe Zhou, Chenhao Xue, Bingzhe Wu, Zhikai Li, Qingyi Gu, Yong Jae Lee, Yan Yan, Beidi Chen, Guangyu Sun, Kurt Keutzer | 2024-05-01T20:42:28Z | cs.CL, cs.AI | The field of efficient Large Language Model (LLM) inference is rapidly
evolving, presenting a unique blend of opportunities and challenges. Although
the field has expanded and is vibrant, there hasn't been a concise framework
that analyzes the various methods of LLM Inference to provide a clear
understanding of this domain. Our survey stands out from traditional literature
reviews by not only summarizing the current state of research but also by
introducing a framework based on roofline model for systematic analysis of LLM
inference techniques. This framework identifies the bottlenecks when deploying
LLMs on hardware devices and provides a clear understanding of practical
problems, such as why LLMs are memory-bound, how much memory and computation
they need, and how to choose the right hardware. We systematically collate the
latest advancements in efficient LLM inference, covering crucial areas such as
model compression (e.g., Knowledge Distillation and Quantization), algorithm
improvements (e.g., Early Exit and Mixture-of-Expert), and both hardware and
system-level enhancements. Our survey stands out by analyzing these methods
with roofline model, helping us understand their impact on memory access and
computation. This distinctive approach not only showcases the current research
landscape but also delivers valuable insights for practical implementation,
positioning our work as an indispensable resource for researchers new to the
field as well as for those seeking to deepen their understanding of efficient
LLM deployment. The analyze tool, LLM-Viewer, is open-sourced. | cs.CL | LLM Inference Unveiled: Survey and Roofline Model Insights |
2024-02-26T07:31:35Z | http://arxiv.org/pdf/2402.16361v1 | http://arxiv.org/abs/2402.16361v1 | Shiwen Ni, Min Yang, Ruifeng Xu, Chengming Li, Xiping Hu | 2024-02-26T07:31:35Z | cs.CL, cs.AI | Among the various pre-trained neural language models that are popular today,
dropout is already an indispensable regularization technique. To solve the
inconsistency between training and inference caused by the randomness of
dropout, some studies use consistency training to regularize dropout at the
output layer. In this paper, we propose a novel Layer-wise Regularized Dropout
(LR-Drop), which is specially designed for Transformer-based Language models.
Specifically, LR-Drop layer-wise regularizes each Transformer layer using the
consistency training strategy. Each training sample passes through the two
siamese sub-models sampled by dropout, and then LR-Drop forces the hidden
states, multi-head attention matrices, and output distribution of the two
siamese sub-models to be consistent. The proposed LR-Drop can be regarded as a
"self-distillation" framework, in which each sub-model generated by dropout is
the other's "teacher" model and "student" model. Through extensive experiments
on 8 natural language understanding datasets, 6 neural machine translation
datasets, and 1 abstractive summarization dataset (a total of 15 datasets), we
show that LR-Drop achieves superior performances, including state-of-the-art
results. | cs.CL | Layer-wise Regularized Dropout for Neural Language Models |
2024-02-26T07:22:51Z | http://arxiv.org/pdf/2402.16358v2 | http://arxiv.org/abs/2402.16358v2 | Yiding Sun, Feng Wang, Yutao Zhu, Wayne Xin Zhao, Jiaxin Mao | 2024-04-23T06:18:32Z | cs.LG, cs.CL, cs.IR | The ability of the foundation models heavily relies on large-scale, diverse,
and high-quality pretraining data. In order to improve data quality,
researchers and practitioners often have to manually curate datasets from
difference sources and develop dedicated data cleansing pipeline for each data
repository. Lacking a unified data processing framework, this process is
repetitive and cumbersome. To mitigate this issue, we propose a data processing
framework that integrates a Processing Module which consists of a series of
operators at different granularity levels, and an Analyzing Module which
supports probing and evaluation of the refined data. The proposed framework is
easy to use and highly flexible. In this demo paper, we first introduce how to
use this framework with some example use cases and then demonstrate its
effectiveness in improving the data quality with an automated evaluation with
ChatGPT and an end-to-end evaluation in pretraining the GPT-2 model. The code
and demonstration videos are accessible on GitHub. | cs.LG | An Integrated Data Processing Framework for Pretraining Foundation Models |
2024-02-26T07:19:23Z | http://arxiv.org/pdf/2402.16354v2 | http://arxiv.org/abs/2402.16354v2 | Haotian Fu, Pratyusha Sharma, Elias Stengel-Eskin, George Konidaris, Nicolas Le Roux, Marc-Alexandre Côté, Xingdi Yuan | 2024-05-27T14:31:38Z | cs.LG, cs.AI, cs.CL | We present an algorithm for skill discovery from expert demonstrations. The
algorithm first utilizes Large Language Models (LLMs) to propose an initial
segmentation of the trajectories. Following that, a hierarchical variational
inference framework incorporates the LLM-generated segmentation information to
discover reusable skills by merging trajectory segments. To further control the
trade-off between compression and reusability, we introduce a novel auxiliary
objective based on the Minimum Description Length principle that helps guide
this skill discovery process. Our results demonstrate that agents equipped with
our method are able to discover skills that help accelerate learning and
outperform baseline skill learning approaches on new long-horizon tasks in
BabyAI, a grid world navigation environment, as well as ALFRED, a household
simulation environment. | cs.LG | Language-guided Skill Learning with Temporal Variational Inference |
2024-02-26T07:17:25Z | http://arxiv.org/pdf/2402.16352v1 | http://arxiv.org/abs/2402.16352v1 | Zimu Lu, Aojun Zhou, Houxing Ren, Ke Wang, Weikang Shi, Junting Pan, Mingjie Zhan, Hongsheng Li | 2024-02-26T07:17:25Z | cs.CL, cs.AI | Large language models (LLMs) have exhibited great potential in mathematical
reasoning. However, there remains a performance gap in this area between
existing open-source models and closed-source models such as GPT-4. In this
paper, we introduce MathGenie, a novel method for generating diverse and
reliable math problems from a small-scale problem-solution dataset (denoted as
seed data). We augment the ground-truth solutions of our seed data and train a
back-translation model to translate the augmented solutions back into new
questions. Subsequently, we generate code-integrated solutions for the new
questions. To ensure the correctness of the code-integrated solutions, we
employ rationale-based strategy for solution verification. Various pretrained
models, ranging from 7B to 70B, are trained on the newly curated data to test
the effectiveness of the proposed augmentation technique, resulting in a family
of models known as MathGenieLM. These models consistently outperform previous
open-source models across five representative mathematical reasoning datasets,
achieving state-of-the-art performance. In particular, MathGenieLM-InternLM2
achieves an accuracy of 87.7% on GSM8K and 55.7% on MATH, securing the best
overall score among open-source language models. | cs.CL | MathGenie: Generating Synthetic Data with Question Back-translation for Enhancing Mathematical Reasoning of LLMs |
2024-02-26T07:00:58Z | http://arxiv.org/pdf/2402.16347v1 | http://arxiv.org/abs/2402.16347v1 | Haoyang Li, Jing Zhang, Hanbing Liu, Ju Fan, Xiaokang Zhang, Jun Zhu, Renjie Wei, Hongyan Pan, Cuiping Li, Hong Chen | 2024-02-26T07:00:58Z | cs.CL, cs.DB | Language models have shown promising performance on the task of translating
natural language questions into SQL queries (Text-to-SQL). However, most of the
state-of-the-art (SOTA) approaches rely on powerful yet closed-source large
language models (LLMs), such as ChatGPT and GPT-4, which may have the
limitations of unclear model architectures, data privacy risks, and expensive
inference overheads. To address the limitations, we introduce CodeS, a series
of pre-trained language models with parameters ranging from 1B to 15B,
specifically designed for the text-to-SQL task. CodeS is a fully open-source
language model, which achieves superior accuracy with much smaller parameter
sizes. This paper studies the research challenges in building CodeS. To enhance
the SQL generation abilities of CodeS, we adopt an incremental pre-training
approach using a specifically curated SQL-centric corpus. Based on this, we
address the challenges of schema linking and rapid domain adaptation through
strategic prompt construction and a bi-directional data augmentation technique.
We conduct comprehensive evaluations on multiple datasets, including the widely
used Spider benchmark, the newly released BIRD benchmark, robustness-diagnostic
benchmarks such as Spider-DK, Spider-Syn, Spider-Realistic, and Dr.Spider, as
well as two real-world datasets created for financial and academic
applications. The experimental results show that our CodeS achieves new SOTA
accuracy and robustness on nearly all challenging text-to-SQL benchmarks. | cs.CL | CodeS: Towards Building Open-source Language Models for Text-to-SQL |
2024-02-26T06:28:54Z | http://arxiv.org/pdf/2402.16333v2 | http://arxiv.org/abs/2402.16333v2 | Xinyi Mou, Zhongyu Wei, Xuanjing Huang | 2024-06-17T05:37:35Z | cs.CY, cs.CL | Social media has emerged as a cornerstone of social movements, wielding
significant influence in driving societal change. Simulating the response of
the public and forecasting the potential impact has become increasingly
important. However, existing methods for simulating such phenomena encounter
challenges concerning their efficacy and efficiency in capturing the behaviors
of social movement participants. In this paper, we introduce a hybrid framework
HiSim for social media user simulation, wherein users are categorized into two
types. Core users are driven by Large Language Models, while numerous ordinary
users are modeled by deductive agent-based models. We further construct a
Twitter-like environment to replicate their response dynamics following trigger
events. Subsequently, we develop a multi-faceted benchmark SoMoSiMu-Bench for
evaluation and conduct comprehensive experiments across real-world datasets.
Experimental results demonstrate the effectiveness and flexibility of our
method. | cs.CY | Unveiling the Truth and Facilitating Change: Towards Agent-based Large-scale Social Movement Simulation |
2024-02-26T05:51:47Z | http://arxiv.org/pdf/2402.16319v1 | http://arxiv.org/abs/2402.16319v1 | Runyu Peng, Yunhua Zhou, Qipeng Guo, Yang Gao, Hang Yan, Xipeng Qiu, Dahua Lin | 2024-02-26T05:51:47Z | cs.CL | Large Language Models (LLMs) are reshaping the research landscape in
artificial intelligence, particularly as model parameters scale up
significantly, unlocking remarkable capabilities across various domains.
Nevertheless, the scalability of model parameters faces constraints due to
limitations in GPU memory and computational speed. To address these
constraints, various weight compression methods have emerged, such as Pruning
and Quantization. Given the low-rank nature of weight matrices in language
models, the reduction of weights through matrix decomposition undoubtedly holds
significant potential and promise. In this paper, drawing upon the intrinsic
structure of LLMs, we propose a novel approach termed Data-free Joint Rank-k
Approximation for compressing the parameter matrices. Significantly, our method
is characterized by without necessitating additional involvement of any corpus,
while simultaneously preserving orthogonality in conjunction with pruning and
quantization methods. We achieve a model pruning of 80% parameters while
retaining 93.43% of the original performance without any calibration data.
Additionally, we explore the fundamental properties of the weight matrix of
LLMs undergone Rank-k Approximation and conduct comprehensive experiments to
elucidate our hypothesis. | cs.CL | Data-freeWeight Compress and Denoise for Large Language Models |
2024-02-26T05:43:51Z | http://arxiv.org/pdf/2402.16315v2 | http://arxiv.org/abs/2402.16315v2 | Jeonghwan Kim, Heng Ji | 2024-03-11T18:49:40Z | cs.CV, cs.CL | Recent advances in instruction-tuned Large Vision-Language Models (LVLMs)
have imbued the models with the ability to generate high-level, image-grounded
explanations with ease. While such capability is largely attributed to the rich
world knowledge contained within the Large Language Models (LLMs), our work
reveals their shortcomings in fine-grained visual categorization (FGVC) across
six different benchmark settings. Most recent state-of-the-art LVLMs like
LLaVa-1.5, InstructBLIP and GPT-4V not only severely deteriorate in terms of
classification performance, e.g., average drop of 65.58 in EM for Stanford Dogs
for LLaVA-1.5, but also struggle to generate an accurate explanation with
detailed attributes based on the concept that appears within an input image
despite their capability to generate holistic image-level descriptions.
In-depth analyses show that instruction-tuned LVLMs exhibit modality gap,
showing discrepancy when given textual and visual inputs that correspond to the
same concept, preventing the image modality from leveraging the rich parametric
knowledge within the LLMs. In an effort to further the community's endeavor in
this direction, we propose a multiple granularity attribute-centric evaluation
benchmark, Finer, which aims to establish a ground to evaluate LVLMs'
fine-grained visual comprehension ability and provide significantly improved
explainability. | cs.CV | Finer: Investigating and Enhancing Fine-Grained Visual Concept Recognition in Large Vision Language Models |
2024-02-26T05:34:16Z | http://arxiv.org/pdf/2403.04771v1 | http://arxiv.org/abs/2403.04771v1 | Lin Ai, Zheng Hui, Zizhou Liu, Julia Hirschberg | 2024-02-26T05:34:16Z | cs.CL | To address the challenges of out-of-control generation in generative models
for machine reading comprehension (MRC), we introduce the Question-Attended
Span Extraction (QASE) module. Integrated during the fine-tuning of pre-trained
generative language models (PLMs), QASE enables these PLMs to match SOTA
extractive methods and outperform leading LLMs like GPT-4 in MRC tasks, without
significant increases in computational costs. | cs.CL | QASE Enhanced PLMs: Improved Control in Text Generation for MRC |
2024-02-26T05:31:34Z | http://arxiv.org/pdf/2402.16313v1 | http://arxiv.org/abs/2402.16313v1 | Mingxu Tao, Dongyan Zhao, Yansong Feng | 2024-02-26T05:31:34Z | cs.CL, cs.AI | Open-ended question answering requires models to find appropriate evidence to
form well-reasoned, comprehensive and helpful answers. In practical
applications, models also need to engage in extended discussions on potential
scenarios closely relevant to the question. With augmentation of retrieval
module, open-source Large Language Models (LLMs) can produce coherent answers
often with different focuses, but are still sub-optimal in terms of reliable
evidence selection and in-depth question analysis. In this paper, we propose a
novel Chain-of-Discussion framework to leverage the synergy among multiple
open-source LLMs aiming to provide \textbf{more correct} and \textbf{more
comprehensive} answers for open-ended QA, although they are not strong enough
individually. Our experiments show that discussions among multiple LLMs play a
vital role in enhancing the quality of answers. We release our data and code at
\url{https://github.com/kobayashikanna01/Chain-of-Discussion}. | cs.CL | Chain-of-Discussion: A Multi-Model Framework for Complex Evidence-Based Question Answering |
2024-02-26T05:30:48Z | http://arxiv.org/pdf/2402.16311v3 | http://arxiv.org/abs/2402.16311v3 | Jingsi Yu, Cunliang Kong, Liner Yang, Meishan Zhang, Lin Zhu, Yujie Wang, Haozhe Lin, Maosong Sun, Erhong Yang | 2024-04-08T03:22:41Z | cs.CL, cs.AI | Sentence Pattern Structure (SPS) parsing is a syntactic analysis method
primarily employed in language teaching.Existing SPS parsers rely heavily on
textbook corpora for training, lacking cross-domain capability.To overcome this
constraint, this paper proposes an innovative approach leveraging large
language models (LLMs) within a self-training framework. Partial syntactic
rules from a source domain are combined with target domain sentences to
dynamically generate training data, enhancing the adaptability of the parser to
diverse domains.Experiments conducted on textbook and news domains demonstrate
the effectiveness of the proposed method, outperforming rule-based baselines by
1.68 points on F1 metrics. | cs.CL | Cross-domain Chinese Sentence Pattern Parsing |
2024-02-26T04:09:53Z | http://arxiv.org/pdf/2402.16288v1 | http://arxiv.org/abs/2402.16288v1 | Yiming Du, Hongru Wang, Zhengyi Zhao, Bin Liang, Baojun Wang, Wanjun Zhong, Zezhong Wang, Kam-Fai Wong | 2024-02-26T04:09:53Z | cs.CL, cs.AI, cs.IR | Long-term memory plays a critical role in personal interaction, considering
long-term memory can better leverage world knowledge, historical information,
and preferences in dialogues. Our research introduces PerLTQA, an innovative QA
dataset that combines semantic and episodic memories, including world
knowledge, profiles, social relationships, events, and dialogues. This dataset
is collected to investigate the use of personalized memories, focusing on
social interactions and events in the QA task. PerLTQA features two types of
memory and a comprehensive benchmark of 8,593 questions for 30 characters,
facilitating the exploration and application of personalized memories in Large
Language Models (LLMs). Based on PerLTQA, we propose a novel framework for
memory integration and generation, consisting of three main components: Memory
Classification, Memory Retrieval, and Memory Synthesis. We evaluate this
framework using five LLMs and three retrievers. Experimental results
demonstrate that BERT-based classification models significantly outperform LLMs
such as ChatGLM3 and ChatGPT in the memory classification task. Furthermore,
our study highlights the importance of effective memory integration in the QA
task. | cs.CL | PerLTQA: A Personal Long-Term Memory Dataset for Memory Classification, Retrieval, and Synthesis in Question Answering |
2024-02-26T03:46:01Z | http://arxiv.org/pdf/2402.16278v3 | http://arxiv.org/abs/2402.16278v3 | Yukihiro Shiraishi, Ken Kaneiwa | 2024-03-10T10:04:41Z | cs.AI, cs.CL, cs.LG | Recently, ontology embeddings representing entities in a low-dimensional
space have been proposed for ontology completion. However, the ontology
embeddings for concept subsumption prediction do not address the difficulties
of similar and isolated entities and fail to extract the global information of
annotation axioms from an ontology. In this paper, we propose a self-matching
training method for the two ontology embedding models: Inverted-index Matrix
Embedding (InME) and Co-occurrence Matrix Embedding (CoME). The two embeddings
capture the global and local information in annotation axioms by means of the
occurring locations of each word in a set of axioms and the co-occurrences of
words in each axiom. The self-matching training method increases the robustness
of the concept subsumption prediction when predicted superclasses are similar
to subclasses and are isolated to other entities in an ontology. Our evaluation
experiments show that the self-matching training method with InME outperforms
the existing ontology embeddings for the GO and FoodOn ontologies and that the
method with the concatenation of CoME and OWL2Vec* outperforms them for the
HeLiS ontology. | cs.AI | A Self-matching Training Method with Annotation Embedding Models for Ontology Subsumption Prediction |
2024-02-26T02:48:43Z | http://arxiv.org/pdf/2402.16261v2 | http://arxiv.org/abs/2402.16261v2 | Hongru Wang, Boyang Xue, Baohang Zhou, Rui Wang, Fei Mi, Weichao Wang, Yasheng Wang, Kam-Fai Wong | 2024-02-28T06:43:48Z | cs.CL, cs.IR | Conversational retrieval refers to an information retrieval system that
operates in an iterative and interactive manner, requiring the retrieval of
various external resources, such as persona, knowledge, and even response, to
effectively engage with the user and successfully complete the dialogue.
However, most previous work trained independent retrievers for each specific
resource, resulting in sub-optimal performance and low efficiency. Thus, we
propose a multi-task framework function as a universal retriever for three
dominant retrieval tasks during the conversation: persona selection, knowledge
selection, and response selection. To this end, we design a dual-encoder
architecture consisting of a context-adaptive dialogue encoder and a candidate
encoder, aiming to attention to the relevant context from the long dialogue and
retrieve suitable candidates by simply a dot product. Furthermore, we introduce
two loss constraints to capture the subtle relationship between dialogue
context and different candidates by regarding historically selected candidates
as hard negatives. Extensive experiments and analysis establish
state-of-the-art retrieval quality both within and outside its training domain,
revealing the promising potential and generalization capability of our model to
serve as a universal retriever for different candidate selection tasks
simultaneously. | cs.CL | UniRetriever: Multi-task Candidates Selection for Various Context-Adaptive Conversational Retrieval |
2024-02-26T02:14:42Z | http://arxiv.org/pdf/2402.16248v1 | http://arxiv.org/abs/2402.16248v1 | Jieyong Wang, Chunyao Song, Yihao Wu | 2024-02-26T02:14:42Z | cs.CL, cs.AI | The topic-to-essay generation task is a challenging natural language
generation task that aims to generate paragraph-level text with high semantic
coherence based on a given set of topic words. Previous work has focused on the
introduction of external knowledge, ignoring the insufficient generated text
diversity. In order to improve the generation diversity, we propose a novel
copy mechanism model with a content selection module that integrates rich
semantic knowledge from the language model into the decoder. Furthermore, we
introduce the improved prefix tuning method to train the model, enabling it to
adapt to varying input complexities. In addition, we have contributed a new
Chinese dataset for TEG tasks. Experimental results demonstrate that the
proposed model can improve the generated text diversity by 35\% to 59\%
compared to the state-of-the-art method, while maintaining a high level of
topic consistency. | cs.CL | Topic-to-essay generation with knowledge-based content selection |
2024-02-26T02:13:36Z | http://arxiv.org/pdf/2402.16247v1 | http://arxiv.org/abs/2402.16247v1 | Dylan Cope, Peter McBurney | 2024-02-26T02:13:36Z | cs.LG, cs.CL, cs.MA | In Emergent Communication (EC) agents learn to communicate with one another,
but the protocols that they develop are specialised to their training
community. This observation led to research into Zero-Shot Coordination (ZSC)
for learning communication strategies that are robust to agents not encountered
during training. However, ZSC typically assumes that no prior data is available
about the agents that will be encountered in the zero-shot setting. In many
cases, this presents an unnecessarily hard problem and rules out communication
via preestablished conventions. We propose a novel AI challenge called a
Cooperative Language Acquisition Problem (CLAP) in which the ZSC assumptions
are relaxed by allowing a 'joiner' agent to learn from a dataset of
interactions between agents in a target community. We propose and compare two
methods for solving CLAPs: Imitation Learning (IL), and Emergent Communication
pretraining and Translation Learning (ECTL), in which an agent is trained in
self-play with EC and then learns from the data to translate between the
emergent protocol and the target community's protocol. | cs.LG | Learning Translations: Emergent Communication Pretraining for Cooperative Language Acquisition |
2024-02-26T01:55:45Z | http://arxiv.org/pdf/2403.04770v1 | http://arxiv.org/abs/2403.04770v1 | Todd Morrill, Zhaoyuan Deng, Yanda Chen, Amith Ananthram, Colin Wayne Leach, Kathleen McKeown | 2024-02-26T01:55:45Z | cs.CL, cs.LG | There are many settings where it is useful to predict and explain the success
or failure of a dialogue. Circumplex theory from psychology models the social
orientations (e.g., Warm-Agreeable, Arrogant-Calculating) of conversation
participants and can be used to predict and explain the outcome of social
interactions. Our work is novel in its systematic application of social
orientation tags to modeling conversation outcomes. In this paper, we introduce
a new data set of dialogue utterances machine-labeled with social orientation
tags. We show that social orientation tags improve task performance, especially
in low-resource settings, on both English and Chinese language benchmarks. We
also demonstrate how social orientation tags help explain the outcomes of
social interactions when used in neural models. Based on these results showing
the utility of social orientation tags for dialogue outcome prediction tasks,
we release our data sets, code, and models that are fine-tuned to predict
social orientation tags on dialogue utterances. | cs.CL | Social Orientation: A New Feature for Dialogue Analysis |
2024-02-25T23:10:20Z | http://arxiv.org/pdf/2403.00815v2 | http://arxiv.org/abs/2403.00815v2 | Ran Xu, Wenqi Shi, Yue Yu, Yuchen Zhuang, Bowen Jin, May D. Wang, Joyce C. Ho, Carl Yang | 2024-06-04T05:11:19Z | cs.CL, cs.AI, cs.IR, q-bio.OT | We present RAM-EHR, a Retrieval AugMentation pipeline to improve clinical
predictions on Electronic Health Records (EHRs). RAM-EHR first collects
multiple knowledge sources, converts them into text format, and uses dense
retrieval to obtain information related to medical concepts. This strategy
addresses the difficulties associated with complex names for the concepts.
RAM-EHR then augments the local EHR predictive model co-trained with
consistency regularization to capture complementary information from patient
visits and summarized knowledge. Experiments on two EHR datasets show the
efficacy of RAM-EHR over previous knowledge-enhanced baselines (3.4% gain in
AUROC and 7.2% gain in AUPR), emphasizing the effectiveness of the summarized
knowledge from RAM-EHR for clinical prediction tasks. The code will be
published at \url{https://github.com/ritaranx/RAM-EHR}. | cs.CL | RAM-EHR: Retrieval Augmentation Meets Clinical Predictions on Electronic Health Records |
2024-02-25T22:23:37Z | http://arxiv.org/pdf/2402.16211v1 | http://arxiv.org/abs/2402.16211v1 | Cem Uluoglakci, Tugba Taskaya Temizel | 2024-02-25T22:23:37Z | cs.CL | Hallucinations pose a significant challenge to the reliability and alignment
of Large Language Models (LLMs), limiting their widespread acceptance beyond
chatbot applications. Despite ongoing efforts, hallucinations remain a
prevalent challenge in LLMs. The detection of hallucinations itself is also a
formidable task, frequently requiring manual labeling or constrained
evaluations. This paper introduces an automated scalable framework that
combines benchmarking LLMs' hallucination tendencies with efficient
hallucination detection. We leverage LLMs to generate challenging tasks related
to hypothetical phenomena, subsequently employing them as agents for efficient
hallucination detection. The framework is domain-agnostic, allowing the use of
any language model for benchmark creation or evaluation in any domain. We
introduce the publicly available HypoTermQA Benchmarking Dataset, on which
state-of-the-art models' performance ranged between 3% and 11%, and evaluator
agents demonstrated a 6% error rate in hallucination prediction. The proposed
framework provides opportunities to test and improve LLMs. Additionally, it has
the potential to generate benchmarking datasets tailored to specific domains,
such as law, health, and finance. | cs.CL | HypoTermQA: Hypothetical Terms Dataset for Benchmarking Hallucination Tendency of LLMs |
2024-02-25T21:25:06Z | http://arxiv.org/pdf/2402.16200v1 | http://arxiv.org/abs/2402.16200v1 | Jianyou Wang, Kaicheng Wang, Xiaoyue Wang, Weili Cao, Ramamohan Paturi, Leon Bergen | 2024-02-25T21:25:06Z | cs.IR, cs.AI, cs.CL, cs.LG | Effective information retrieval (IR) in settings with limited training data,
particularly for complex queries, remains a challenging task. This paper
introduces IR2, Information Regularization for Information Retrieval, a
technique for reducing overfitting during synthetic data generation. This
approach, representing a novel application of regularization techniques in
synthetic data creation for IR, is tested on three recent IR tasks
characterized by complex queries: DORIS-MAE, ArguAna, and WhatsThatBook.
Experimental results indicate that our regularization techniques not only
outperform previous synthetic query generation methods on the tasks considered
but also reduce cost by up to 50%. Furthermore, this paper categorizes and
explores three regularization methods at different stages of the query
synthesis pipeline-input, prompt, and output-each offering varying degrees of
performance improvement compared to models where no regularization is applied.
This provides a systematic approach for optimizing synthetic data generation in
data-limited, complex-query IR scenarios. All code, prompts and synthetic data
are available at
https://github.com/Info-Regularization/Information-Regularization. | cs.IR | IR2: Information Regularization for Information Retrieval |
2024-02-25T20:36:51Z | http://arxiv.org/pdf/2402.16194v1 | http://arxiv.org/abs/2402.16194v1 | Omama Hamad, Ali Hamdi, Khaled Shaban | 2024-02-25T20:36:51Z | cs.CL | Effective feature representations play a critical role in enhancing the
performance of text generation models that rely on deep neural networks.
However, current approaches suffer from several drawbacks, such as the
inability to capture the deep semantics of language and sensitivity to minor
input variations, resulting in significant changes in the generated text. In
this paper, we present a novel solution to these challenges by employing a
mixture of experts, multiple encoders, to offer distinct perspectives on the
emotional state of the user's utterance while simultaneously enhancing
performance. We propose an end-to-end model architecture called ASEM that
performs emotion analysis on top of sentiment analysis for open-domain
chatbots, enabling the generation of empathetic responses that are fluent and
relevant. In contrast to traditional attention mechanisms, the proposed model
employs a specialized attention strategy that uniquely zeroes in on sentiment
and emotion nuances within the user's utterance. This ensures the generation of
context-rich representations tailored to the underlying emotional tone and
sentiment intricacies of the text. Our approach outperforms existing methods
for generating empathetic embeddings, providing empathetic and diverse
responses. The performance of our proposed model significantly exceeds that of
existing models, enhancing emotion detection accuracy by 6.2% and lexical
diversity by 1.4%. | cs.CL | ASEM: Enhancing Empathy in Chatbot through Attention-based Sentiment and Emotion Modeling |
2024-02-25T20:36:03Z | http://arxiv.org/pdf/2402.16192v2 | http://arxiv.org/abs/2402.16192v2 | Jiabao Ji, Bairu Hou, Alexander Robey, George J. Pappas, Hamed Hassani, Yang Zhang, Eric Wong, Shiyu Chang | 2024-02-28T23:11:33Z | cs.CL | Aligned large language models (LLMs) are vulnerable to jailbreaking attacks,
which bypass the safeguards of targeted LLMs and fool them into generating
objectionable content. While initial defenses show promise against token-based
threat models, there do not exist defenses that provide robustness against
semantic attacks and avoid unfavorable trade-offs between robustness and
nominal performance. To meet this need, we propose SEMANTICSMOOTH, a
smoothing-based defense that aggregates the predictions of multiple
semantically transformed copies of a given input prompt. Experimental results
demonstrate that SEMANTICSMOOTH achieves state-of-the-art robustness against
GCG, PAIR, and AutoDAN attacks while maintaining strong nominal performance on
instruction following benchmarks such as InstructionFollowing and AlpacaEval.
The codes will be publicly available at
https://github.com/UCSB-NLP-Chang/SemanticSmooth. | cs.CL | Defending Large Language Models against Jailbreak Attacks via Semantic Smoothing |
2024-02-25T20:24:07Z | http://arxiv.org/pdf/2402.16187v2 | http://arxiv.org/abs/2402.16187v2 | Qi Pang, Shengyuan Hu, Wenting Zheng, Virginia Smith | 2024-05-25T20:51:49Z | cs.CR, cs.CL, cs.LG | Advances in generative models have made it possible for AI-generated text,
code, and images to mirror human-generated content in many applications.
Watermarking, a technique that aims to embed information in the output of a
model to verify its source, is useful for mitigating the misuse of such
AI-generated content. However, we show that common design choices in LLM
watermarking schemes make the resulting systems surprisingly susceptible to
attack -- leading to fundamental trade-offs in robustness, utility, and
usability. To navigate these trade-offs, we rigorously study a set of simple
yet effective attacks on common watermarking systems, and propose guidelines
and defenses for LLM watermarking in practice. | cs.CR | No Free Lunch in LLM Watermarking: Trade-offs in Watermarking Design Choices |
2024-02-25T18:33:25Z | http://arxiv.org/pdf/2402.16168v1 | http://arxiv.org/abs/2402.16168v1 | Avik Pal, Madhura Pawar | 2024-02-25T18:33:25Z | cs.CL, cs.AI | Structural probes learn a linear transformation to find how dependency trees
are embedded in the hidden states of language models. This simple design may
not allow for full exploitation of the structure of the encoded information.
Hence, to investigate the structure of the encoded information to its full
extent, we incorporate non-linear structural probes. We reformulate the design
of non-linear structural probes introduced by White et al. making its design
simpler yet effective. We also design a visualization framework that lets us
qualitatively assess how strongly two words in a sentence are connected in the
predicted dependency tree. We use this technique to understand which non-linear
probe variant is good at encoding syntactical information. Additionally, we
also use it to qualitatively investigate the structure of dependency trees that
BERT encodes in each of its layers. We find that the radial basis function
(RBF) is an effective non-linear probe for the BERT model than the linear
probe. | cs.CL | Hitting "Probe"rty with Non-Linearity, and More |
2024-02-25T17:43:29Z | http://arxiv.org/pdf/2402.16914v2 | http://arxiv.org/abs/2402.16914v2 | Xirui Li, Ruochen Wang, Minhao Cheng, Tianyi Zhou, Cho-Jui Hsieh | 2024-03-01T07:26:50Z | cs.CR, cs.AI, cs.CL | The safety alignment of Large Language Models (LLMs) is vulnerable to both
manual and automated jailbreak attacks, which adversarially trigger LLMs to
output harmful content. However, current methods for jailbreaking LLMs, which
nest entire harmful prompts, are not effective at concealing malicious intent
and can be easily identified and rejected by well-aligned LLMs. This paper
discovers that decomposing a malicious prompt into separated sub-prompts can
effectively obscure its underlying malicious intent by presenting it in a
fragmented, less detectable form, thereby addressing these limitations. We
introduce an automatic prompt \textbf{D}ecomposition and
\textbf{R}econstruction framework for jailbreak \textbf{Attack} (DrAttack).
DrAttack includes three key components: (a) `Decomposition' of the original
prompt into sub-prompts, (b) `Reconstruction' of these sub-prompts implicitly
by in-context learning with semantically similar but harmless reassembling
demo, and (c) a `Synonym Search' of sub-prompts, aiming to find sub-prompts'
synonyms that maintain the original intent while jailbreaking LLMs. An
extensive empirical study across multiple open-source and closed-source LLMs
demonstrates that, with a significantly reduced number of queries, DrAttack
obtains a substantial gain of success rate over prior SOTA prompt-only
attackers. Notably, the success rate of 78.0\% on GPT-4 with merely 15 queries
surpassed previous art by 33.1\%. The project is available at
https://github.com/xirui-li/DrAttack. | cs.CR | DrAttack: Prompt Decomposition and Reconstruction Makes Powerful LLM Jailbreakers |
2024-02-25T17:40:49Z | http://arxiv.org/pdf/2402.16159v5 | http://arxiv.org/abs/2402.16159v5 | Somnath Banerjee, Avik Dutta, Aaditya Agrawal, Rima Hazra, Animesh Mukherjee | 2024-06-20T12:43:57Z | cs.CL | With the AI revolution in place, the trend for building automated systems to
support professionals in different domains such as the open source software
systems, healthcare systems, banking systems, transportation systems and many
others have become increasingly prominent. A crucial requirement in the
automation of support tools for such systems is the early identification of
named entities, which serves as a foundation for developing specialized
functionalities. However, due to the specific nature of each domain, different
technical terminologies and specialized languages, expert annotation of
available data becomes expensive and challenging. In light of these challenges,
this paper proposes a novel named entity recognition (NER) technique
specifically tailored for the open-source software systems. Our approach aims
to address the scarcity of annotated software data by employing a comprehensive
two-step distantly supervised annotation process. This process strategically
leverages language heuristics, unique lookup tables, external knowledge
sources, and an active learning approach. By harnessing these powerful
techniques, we not only enhance model performance but also effectively mitigate
the limitations associated with cost and the scarcity of expert annotators. It
is noteworthy that our model significantly outperforms the state-of-the-art
LLMs by a substantial margin. We also show the effectiveness of NER in the
downstream task of relation extraction. | cs.CL | DistALANER: Distantly Supervised Active Learning Augmented Named Entity Recognition in the Open Source Software Ecosystem |
2024-02-25T17:19:41Z | http://arxiv.org/pdf/2402.16153v1 | http://arxiv.org/abs/2402.16153v1 | Ruibin Yuan, Hanfeng Lin, Yi Wang, Zeyue Tian, Shangda Wu, Tianhao Shen, Ge Zhang, Yuhang Wu, Cong Liu, Ziya Zhou, Ziyang Ma, Liumeng Xue, Ziyu Wang, Qin Liu, Tianyu Zheng, Yizhi Li, Yinghao Ma, Yiming Liang, Xiaowei Chi, Ruibo Liu, Zili Wang, Pengfei Li, Jingcheng Wu, Chenghua Lin, Qifeng Liu, Tao Jiang, Wenhao Huang, Wenhu Chen, Emmanouil Benetos, Jie Fu, Gus Xia, Roger Dannenberg, Wei Xue, Shiyin Kang, Yike Guo | 2024-02-25T17:19:41Z | cs.SD, cs.AI, cs.CL, cs.LG, cs.MM, eess.AS | While Large Language Models (LLMs) demonstrate impressive capabilities in
text generation, we find that their ability has yet to be generalized to music,
humanity's creative language. We introduce ChatMusician, an open-source LLM
that integrates intrinsic musical abilities. It is based on continual
pre-training and finetuning LLaMA2 on a text-compatible music representation,
ABC notation, and the music is treated as a second language. ChatMusician can
understand and generate music with a pure text tokenizer without any external
multi-modal neural structures or tokenizers. Interestingly, endowing musical
abilities does not harm language abilities, even achieving a slightly higher
MMLU score. Our model is capable of composing well-structured, full-length
music, conditioned on texts, chords, melodies, motifs, musical forms, etc,
surpassing GPT-4 baseline. On our meticulously curated college-level music
understanding benchmark, MusicTheoryBench, ChatMusician surpasses LLaMA2 and
GPT-3.5 on zero-shot setting by a noticeable margin. Our work reveals that LLMs
can be an excellent compressor for music, but there remains significant
territory to be conquered. We release our 4B token music-language corpora
MusicPile, the collected MusicTheoryBench, code, model and demo in GitHub. | cs.SD | ChatMusician: Understanding and Generating Music Intrinsically with LLM |
2024-02-25T16:47:59Z | http://arxiv.org/pdf/2402.16142v1 | http://arxiv.org/abs/2402.16142v1 | Pravneet Kaur, Gautam Siddharth Kashyap, Ankit Kumar, Md Tabrez Nafis, Sandeep Kumar, Vikrant Shokeen | 2024-02-25T16:47:59Z | cs.CL, cs.AI | This groundbreaking study explores the expanse of Large Language Models
(LLMs), such as Generative Pre-Trained Transformer (GPT) and Bidirectional
Encoder Representations from Transformers (BERT) across varied domains ranging
from technology, finance, healthcare to education. Despite their established
prowess in Natural Language Processing (NLP), these LLMs have not been
systematically examined for their impact on domains such as fitness, and
holistic well-being, urban planning, climate modelling as well as disaster
management. This review paper, in addition to furnishing a comprehensive
analysis of the vast expanse and extent of LLMs' utility in diverse domains,
recognizes the research gaps and realms where the potential of LLMs is yet to
be harnessed. This study uncovers innovative ways in which LLMs can leave a
mark in the fields like fitness and wellbeing, urban planning, climate
modelling and disaster response which could inspire future researches and
applications in the said avenues. | cs.CL | From Text to Transformation: A Comprehensive Review of Large Language Models' Versatility |
2024-02-25T16:43:41Z | http://arxiv.org/pdf/2402.16141v1 | http://arxiv.org/abs/2402.16141v1 | Xiangdi Meng, Damai Dai, Weiyao Luo, Zhe Yang, Shaoxiang Wu, Xiaochen Wang, Peiyi Wang, Qingxiu Dong, Liang Chen, Zhifang Sui | 2024-02-25T16:43:41Z | cs.CL | Supervised fine-tuning is the most common method to adapt large language
models (LLMs) to downstream tasks, but full fine-tuning LLMs requires massive
computational resources. Recently, parameter-efficient fine-tuning (PEFT)
methods have been widely studied due to its cost-effectiveness. LoRA is one of
the most widely used methods, which assumes that the optimization process is
essentially low-dimensional. Although LoRA fine-tuning is effective, there is
still a performance gap compared to full fine-tuning, since its weight update
is limited to low-rank matrices. In order to break the low-rank bottleneck in
LoRA Optimization, we propose PeriodicLoRA (PLoRA), which accumulates low-rank
update matrices multiple times to achieve a higher update rank. PLoRA has
multiple training stages. During each stage, we still update only the LoRA
weights. However, at the end of each stage, we unload the LoRA weights into the
backbone parameters and then reinitialize the LoRA states. Experimental results
show that PLoRA has stronger learning ability, approximately 1.8 times that of
LoRA's learning ability at most, but it does not increase memory usage.
Further, we introduce a momentum-based unloading strategy for PLoRA to mitigate
the training instability. | cs.CL | PeriodicLoRA: Breaking the Low-Rank Bottleneck in LoRA Optimization |
2024-02-25T16:36:51Z | http://arxiv.org/pdf/2402.16139v3 | http://arxiv.org/abs/2402.16139v3 | Antonio San Martín | 2024-04-19T16:13:43Z | cs.CL, cs.AI | This paper examines the impact of Generative Artificial Intelligence (GenAI)
tools like ChatGPT on the creation and consumption of terminological
definitions. From the terminologist's point of view, the strategic use of GenAI
tools can streamline the process of crafting definitions, reducing both time
and effort, while potentially enhancing quality. GenAI tools enable AI-assisted
terminography, notably post-editing terminography, where the machine produces a
definition that the terminologist then corrects or refines. However, the
potential of GenAI tools to fulfill all the terminological needs of a user,
including term definitions, challenges the very existence of terminological
definitions and resources as we know them. Unlike terminological definitions,
GenAI tools can describe the knowledge activated by a term in a specific
context. However, a main drawback of these tools is that their output can
contain errors. For this reason, users requiring reliability will likely still
resort to terminological resources for definitions. Nevertheless, with the
inevitable integration of AI into terminology work, the distinction between
human-created and AI-created content will become increasingly blurred. | cs.CL | What Generative Artificial Intelligence Means for Terminological Definitions |
2024-02-25T16:14:26Z | http://arxiv.org/pdf/2402.16132v1 | http://arxiv.org/abs/2402.16132v1 | Haoxin Liu, Zhiyuan Zhao, Jindong Wang, Harshavardhan Kamarthi, B. Aditya Prakash | 2024-02-25T16:14:26Z | cs.CL, cs.AI | Time-series forecasting (TSF) finds broad applications in real-world
scenarios. Prompting off-the-shelf Large Language Models (LLMs) demonstrates
strong zero-shot TSF capabilities while preserving computational efficiency.
However, existing prompting methods oversimplify TSF as language next-token
predictions, overlooking its dynamic nature and lack of integration with
state-of-the-art prompt strategies such as Chain-of-Thought. Thus, we propose
LSTPrompt, a novel approach for prompting LLMs in zero-shot TSF tasks.
LSTPrompt decomposes TSF into short-term and long-term forecasting sub-tasks,
tailoring prompts to each. LSTPrompt guides LLMs to regularly reassess
forecasting mechanisms to enhance adaptability. Extensive evaluations
demonstrate consistently better performance of LSTPrompt than existing
prompting methods, and competitive results compared to foundation TSF models. | cs.CL | LSTPrompt: Large Language Models as Zero-Shot Time Series Forecasters by Long-Short-Term Prompting |
2024-02-25T15:46:33Z | http://arxiv.org/pdf/2402.16123v2 | http://arxiv.org/abs/2402.16123v2 | Ningyu Zhang, Bozhong Tian, Siyuan Cheng, Xiaozhuan Liang, Yi Hu, Kouying Xue, Yanjie Gou, Xi Chen, Huajun Chen | 2024-04-28T12:03:38Z | cs.CL, cs.AI, cs.CV, cs.HC, cs.LG | Knowledge editing for large language models can offer an efficient solution
to alter a model's behavior without negatively impacting the overall
performance. However, the current approaches encounter issues with limited
generalizability across tasks, necessitating one distinct editor for each task,
significantly hindering the broader applications. To address this, we take the
first step to analyze the multi-task generalization issue in knowledge editing.
Specifically, we develop an instruction-based editing technique, termed
InstructEdit, which facilitates the editor's adaptation to various task
performances simultaneously using simple instructions. With only one unified
editor for each LLM, we empirically demonstrate that InstructEdit can improve
the editor's control, leading to an average 14.86% increase in Reliability in
multi-task editing setting. Furthermore, experiments involving holdout unseen
task illustrate that InstructEdit consistently surpass previous strong
baselines. To further investigate the underlying mechanisms of
instruction-based knowledge editing, we analyze the principal components of the
editing gradient directions, which unveils that instructions can help control
optimization direction with stronger OOD generalization. Code and datasets are
available in https://github.com/zjunlp/EasyEdit. | cs.CL | InstructEdit: Instruction-based Knowledge Editing for Large Language Models |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.