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
|
|---|---|---|---|---|---|---|---|---|
2024-06-17T18:57:37Z | http://arxiv.org/pdf/2406.12030v1 | http://arxiv.org/abs/2406.12030v1 | Yongting Zhang, Lu Chen, Guodong Zheng, Yifeng Gao, Rui Zheng, Jinlan Fu, Zhenfei Yin, Senjie Jin, Yu Qiao, Xuanjing Huang, Feng Zhao, Tao Gui, Jing Shao | 2024-06-17T18:57:37Z | cs.CV, cs.AI, cs.CL | The emergence of Vision Language Models (VLMs) has brought unprecedented
advances in understanding multimodal information. The combination of textual
and visual semantics in VLMs is highly complex and diverse, making the safety
alignment of these models challenging. Furthermore, due to the limited study on
the safety alignment of VLMs, there is a lack of large-scale, high-quality
datasets. To address these limitations, we propose a Safety Preference
Alignment dataset for Vision Language Models named SPA-VL. In terms of breadth,
SPA-VL covers 6 harmfulness domains, 13 categories, and 53 subcategories, and
contains 100,788 samples of the quadruple (question, image, chosen response,
rejected response). In terms of depth, the responses are collected from 12
open- (e.g., QwenVL) and closed-source (e.g., Gemini) VLMs to ensure diversity.
The experimental results indicate that models trained with alignment techniques
on the SPA-VL dataset exhibit substantial improvements in harmlessness and
helpfulness while maintaining core capabilities. SPA-VL, as a large-scale,
high-quality, and diverse dataset, represents a significant milestone in
ensuring that VLMs achieve both harmlessness and helpfulness. We have made our
code https://github.com/EchoseChen/SPA-VL-RLHF and SPA-VL dataset url
https://huggingface.co/datasets/sqrti/SPA-VL publicly available. | cs.CV | SPA-VL: A Comprehensive Safety Preference Alignment Dataset for Vision Language Model |
2024-06-17T18:45:41Z | http://arxiv.org/pdf/2406.12023v1 | http://arxiv.org/abs/2406.12023v1 | Christian Herold, Michael Kozielski, Leonid Ekimov, Pavel Petrushkov, Pierre-Yves Vandenbussche, Shahram Khadivi | 2024-06-17T18:45:41Z | cs.CL, cs.LG | We introduce the LiLiuM series of large language models (LLMs): 1B, 7B, and
13B parameter models developed 100% in-house to fit eBay's specific needs in
the e-commerce domain. This gives eBay full control over all aspects of the
models including license, data, vocabulary, and architecture. We expect these
models to be used as a foundation for fine-tuning and instruction-tuning,
eliminating dependencies to external models.
The LiLiuM LLMs have been trained on 3 trillion tokens of multilingual text
from general and e-commerce domain. They perform similar to the popular LLaMA-2
models on English natural language understanding (NLU) benchmarks. At the same
time, we outperform LLaMA-2 on non-English NLU tasks, machine translation and
on e-commerce specific downstream tasks.
As part of our data mixture, we utilize the newly released RedPajama-V2
dataset for training and share our insights regarding data filtering and
deduplication. We also discuss in detail how to serialize structured data for
use in autoregressive language modeling. We provide insights on the effects of
including code and parallel machine translation data in pre-training.
Furthermore, we develop our own tokenizer and model vocabulary, customized
towards e-commerce. This way, we can achieve up to 34% speed-up in text
generation on eBay-specific downstream tasks compared to LLaMA-2.
Finally, in relation to LLM pretraining, we show that checkpoint averaging
can further improve over the best individual model checkpoint. | cs.CL | LiLiuM: eBay's Large Language Models for e-commerce |
2024-06-17T18:34:58Z | http://arxiv.org/pdf/2406.12018v1 | http://arxiv.org/abs/2406.12018v1 | Yu Bai, Xiyuan Zou, Heyan Huang, Sanxing Chen, Marc-Antoine Rondeau, Yang Gao, Jackie Chi Kit Cheung | 2024-06-17T18:34:58Z | cs.CL | Long sequence modeling has gained broad interest as large language models
(LLMs) continue to advance. Recent research has identified that a large portion
of hidden states within the key-value caches of Transformer models can be
discarded (also termed evicted) without affecting the perplexity performance in
generating long sequences. However, we show that these methods, despite
preserving perplexity performance, often drop information that is important for
solving downstream tasks, a problem which we call information neglect. To
address this issue, we introduce Chunked Instruction-aware State Eviction
(CItruS), a novel modeling technique that integrates the attention preferences
useful for a downstream task into the eviction process of hidden states. In
addition, we design a method for chunked sequence processing to further improve
efficiency. Our training-free method exhibits superior performance on long
sequence comprehension and retrieval tasks over several strong baselines under
the same memory budget, while preserving language modeling perplexity. | cs.CL | CItruS: Chunked Instruction-aware State Eviction for Long Sequence Modeling |
2024-06-17T18:25:02Z | http://arxiv.org/pdf/2406.12009v1 | http://arxiv.org/abs/2406.12009v1 | Ziyue Xu, Peilin Zhou, Xinyu Shi, Jiageng Wu, Yikang Jiang, Bin Ke, Jie Yang | 2024-06-17T18:25:02Z | cs.CL | Accurate and transparent financial information disclosure is crucial in the
fields of accounting and finance, ensuring market efficiency and investor
confidence. Among many information disclosure platforms, the Chinese stock
exchanges' investor interactive platform provides a novel and interactive way
for listed firms to disclose information of interest to investors through an
online question-and-answer (Q&A) format. However, it is common for listed firms
to respond to questions with limited or no substantive information, and
automatically evaluating the quality of financial information disclosure on
large amounts of Q&A pairs is challenging. This paper builds a benchmark
FinTruthQA, that can evaluate advanced natural language processing (NLP)
techniques for the automatic quality assessment of information disclosure in
financial Q&A data. FinTruthQA comprises 6,000 real-world financial Q&A entries
and each Q&A was manually annotated based on four conceptual dimensions of
accounting. We benchmarked various NLP techniques on FinTruthQA, including
statistical machine learning models, pre-trained language model and their
fine-tuned versions, as well as the large language model GPT-4. Experiments
showed that existing NLP models have strong predictive ability for real
question identification and question relevance tasks, but are suboptimal for
answer relevance and answer readability tasks. By establishing this benchmark,
we provide a robust foundation for the automatic evaluation of information
disclosure, significantly enhancing the transparency and quality of financial
reporting. FinTruthQA can be used by auditors, regulators, and financial
analysts for real-time monitoring and data-driven decision-making, as well as
by researchers for advanced studies in accounting and finance, ultimately
fostering greater trust and efficiency in the financial markets. | cs.CL | FinTruthQA: A Benchmark Dataset for Evaluating the Quality of Financial Information Disclosure |
2024-06-17T18:01:32Z | http://arxiv.org/pdf/2406.11978v1 | http://arxiv.org/abs/2406.11978v1 | Kenneth Li, Yiming Wang, Fernanda Viégas, Martin Wattenberg | 2024-06-17T18:01:32Z | cs.CL, cs.AI, cs.LG | We present an approach called Dialogue Action Tokens (DAT) that adapts
language model agents to plan goal-directed dialogues. The core idea is to
treat each utterance as an action, thereby converting dialogues into games
where existing approaches such as reinforcement learning can be applied.
Specifically, we freeze a pretrained language model and train a small planner
model that predicts a continuous action vector, used for controlled generation
in each round. This design avoids the problem of language degradation under
reward optimization. When evaluated on the Sotopia platform for social
simulations, the DAT-steered LLaMA model surpasses GPT-4's performance. We also
apply DAT to steer an attacker language model in a novel multi-turn red-teaming
setting, revealing a potential new attack surface. | cs.CL | Dialogue Action Tokens: Steering Language Models in Goal-Directed Dialogue with a Multi-Turn Planner |
2024-06-17T18:01:06Z | http://arxiv.org/pdf/2406.11977v1 | http://arxiv.org/abs/2406.11977v1 | Eva Portelance, Siva Reddy, Timothy J. O'Donnell | 2024-06-17T18:01:06Z | cs.CL, I.2.7; I.2.10; I.2.6; F.4.2 | Semantic and syntactic bootstrapping posit that children use their prior
knowledge of one linguistic domain, say syntactic relations, to help later
acquire another, such as the meanings of new words. Empirical results
supporting both theories may tempt us to believe that these are different
learning strategies, where one may precede the other. Here, we argue that they
are instead both contingent on a more general learning strategy for language
acquisition: joint learning. Using a series of neural visually-grounded grammar
induction models, we demonstrate that both syntactic and semantic bootstrapping
effects are strongest when syntax and semantics are learnt simultaneously.
Joint learning results in better grammar induction, realistic lexical category
learning, and better interpretations of novel sentence and verb meanings. Joint
learning makes language acquisition easier for learners by mutually
constraining the hypotheses spaces for both syntax and semantics. Studying the
dynamics of joint inference over many input sources and modalities represents
an important new direction for language modeling and learning research in both
cognitive sciences and AI, as it may help us explain how language can be
acquired in more constrained learning settings. | cs.CL | Reframing linguistic bootstrapping as joint inference using visually-grounded grammar induction models |
2024-06-17T17:59:58Z | http://arxiv.org/pdf/2406.11839v1 | http://arxiv.org/abs/2406.11839v1 | Fei Wang, Wenxuan Zhou, James Y. Huang, Nan Xu, Sheng Zhang, Hoifung Poon, Muhao Chen | 2024-06-17T17:59:58Z | cs.CV, cs.AI, cs.CL, cs.LG | Direct preference optimization (DPO) has shown to be an effective method for
large language model (LLM) alignment. Recent works have attempted to apply DPO
to multimodal scenarios but have found it challenging to achieve consistent
improvement. Through a comparative experiment, we identify the unconditional
preference problem in multimodal preference optimization, where the model
overlooks the image condition. To address this problem, we propose mDPO, a
multimodal DPO objective that prevents the over-prioritization of language-only
preferences by also optimizing image preference. Moreover, we introduce a
reward anchor that forces the reward to be positive for chosen responses,
thereby avoiding the decrease in their likelihood -- an intrinsic problem of
relative preference optimization. Experiments on two multimodal LLMs of
different sizes and three widely used benchmarks demonstrate that mDPO
effectively addresses the unconditional preference problem in multimodal
preference optimization and significantly improves model performance,
particularly in reducing hallucination. | cs.CV | mDPO: Conditional Preference Optimization for Multimodal Large Language Models |
2024-06-17T17:59:35Z | http://arxiv.org/pdf/2406.11830v1 | http://arxiv.org/abs/2406.11830v1 | Belinda Z. Li, Emmy Liu, Alexis Ross, Abbas Zeitoun, Graham Neubig, Jacob Andreas | 2024-06-17T17:59:35Z | cs.CL, cs.AI | When the world changes, so does the text that humans write about it. How do
we build language models that can be easily updated to reflect these changes?
One popular approach is retrieval-augmented generation, in which new documents
are inserted into a knowledge base and retrieved during prediction for
downstream tasks. Most prior work on these systems have focused on improving
behavior during prediction through better retrieval or reasoning. This paper
introduces ERASE, which instead improves model behavior when new documents are
acquired, by incrementally deleting or rewriting other entries in the knowledge
base each time a document is added. In two new benchmark datasets evaluating
models' ability to answer questions about a stream of news articles or
conversations, ERASE improves accuracy relative to conventional
retrieval-augmented generation by 7-13% (Mixtral-8x7B) and 6-10% (Llama-3-8B)
absolute. Code and data are available at https://github.com/belindal/ERASE | cs.CL | Language Modeling with Editable External Knowledge |
2024-06-17T17:59:13Z | http://arxiv.org/pdf/2406.11827v1 | http://arxiv.org/abs/2406.11827v1 | Wenxuan Zhou, Ravi Agrawal, Shujian Zhang, Sathish Reddy Indurthi, Sanqiang Zhao, Kaiqiang Song, Silei Xu, Chenguang Zhu | 2024-06-17T17:59:13Z | cs.CL, cs.AI, cs.LG | Reinforcement learning from human feedback (RLHF) is a promising solution to
align large language models (LLMs) more closely with human values. Off-policy
preference optimization, where the preference data is obtained from other
models, is widely adopted due to its cost efficiency and scalability. However,
off-policy preference optimization often suffers from a distributional gap
between the policy used for data collection and the target policy, leading to
suboptimal optimization. In this paper, we propose a novel strategy to mitigate
this problem by simulating on-policy learning with off-policy preference data.
Our Weighted Preference Optimization (WPO) method adapts off-policy data to
resemble on-policy data more closely by reweighting preference pairs according
to their probability under the current policy. This method not only addresses
the distributional gap problem but also enhances the optimization process
without incurring additional costs. We validate our method on instruction
following benchmarks including Alpaca Eval 2 and MT-bench. WPO not only
outperforms Direct Preference Optimization (DPO) by up to 5.6% on Alpaca Eval 2
but also establishes a remarkable length-controlled winning rate against
GPT-4-turbo of 48.6% based on Llama-3-8B-Instruct, making it the strongest 8B
model on the leaderboard. We will release the code and models at
https://github.com/wzhouad/WPO. | cs.CL | WPO: Enhancing RLHF with Weighted Preference Optimization |
2024-06-17T17:57:30Z | http://arxiv.org/pdf/2406.11823v1 | http://arxiv.org/abs/2406.11823v1 | Geewook Kim, Minjoon Seo | 2024-06-17T17:57:30Z | cs.CV, cs.CL | Recent advancements in language and vision assistants have showcased
impressive capabilities but suffer from a lack of transparency, limiting
broader research and reproducibility. While open-source models handle general
image tasks effectively, they face challenges with the high computational
demands of complex visually-situated text understanding. Such tasks often
require increased token inputs and large vision modules to harness
high-resolution information. Striking a balance between model size and data
importance remains an open question. This study aims to redefine the design of
vision-language models by identifying key components and creating efficient
models with constrained inference costs. By strategically formulating datasets,
optimizing vision modules, and enhancing supervision techniques, we achieve
significant improvements in inference throughput while maintaining high
performance. Extensive experiments across models ranging from 160M to 13B
parameters offer insights into model optimization. We will fully open-source
our codebase, models, and datasets at https://github.com/naver-ai/elva . | cs.CV | On Efficient Language and Vision Assistants for Visually-Situated Natural Language Understanding: What Matters in Reading and Reasoning |
2024-06-17T17:55:38Z | http://arxiv.org/pdf/2406.11817v1 | http://arxiv.org/abs/2406.11817v1 | Jie Liu, Zhanhui Zhou, Jiaheng Liu, Xingyuan Bu, Chao Yang, Han-Sen Zhong, Wanli Ouyang | 2024-06-17T17:55:38Z | cs.CL, cs.AI, cs.LG | Direct Preference Optimization (DPO), a standard method for aligning language
models with human preferences, is traditionally applied to offline preferences.
Recent studies show that DPO benefits from iterative training with online
preferences labeled by a trained reward model. In this work, we identify a
pitfall of vanilla iterative DPO - improved response quality can lead to
increased verbosity. To address this, we introduce iterative length-regularized
DPO (iLR-DPO) to penalize response length. Our empirical results show that
iLR-DPO can enhance a 7B model to perform on par with GPT-4 without increasing
verbosity. Specifically, our 7B model achieves a $50.5\%$ length-controlled win
rate against $\texttt{GPT-4 Preview}$ on AlpacaEval 2.0, and excels across
standard benchmarks including MT-Bench, Arena-Hard and OpenLLM Leaderboard.
These results demonstrate the effectiveness of iterative DPO in aligning
language models with human feedback. | cs.CL | Iterative Length-Regularized Direct Preference Optimization: A Case Study on Improving 7B Language Models to GPT-4 Level |
2024-06-17T17:54:40Z | http://arxiv.org/pdf/2406.11813v1 | http://arxiv.org/abs/2406.11813v1 | Hoyeon Chang, Jinho Park, Seonghyeon Ye, Sohee Yang, Youngkyung Seo, Du-Seong Chang, Minjoon Seo | 2024-06-17T17:54:40Z | cs.CL, I.2.7 | Despite the recent observation that large language models (LLMs) can store
substantial factual knowledge, there is a limited understanding of the
mechanisms of how they acquire factual knowledge through pretraining. This work
addresses this gap by studying how LLMs acquire factual knowledge during
pretraining. The findings reveal several important insights into the dynamics
of factual knowledge acquisition during pretraining. First, counterintuitively,
we observe that pretraining on more data shows no significant improvement in
the model's capability to acquire and maintain factual knowledge. Next, there
is a power-law relationship between training steps and forgetting of
memorization and generalization of factual knowledge, and LLMs trained with
duplicated training data exhibit faster forgetting. Third, training LLMs with
larger batch sizes can enhance the models' robustness to forgetting. Overall,
our observations suggest that factual knowledge acquisition in LLM pretraining
occurs by progressively increasing the probability of factual knowledge
presented in the pretraining data at each step. However, this increase is
diluted by subsequent forgetting. Based on this interpretation, we demonstrate
that we can provide plausible explanations for recently observed behaviors of
LLMs, such as the poor performance of LLMs on long-tail knowledge and the
benefits of deduplicating the pretraining corpus. | cs.CL | How Do Large Language Models Acquire Factual Knowledge During Pretraining? |
2024-06-17T17:52:54Z | http://arxiv.org/pdf/2406.11811v1 | http://arxiv.org/abs/2406.11811v1 | Joao Monteiro, Pierre-Andre Noel, Etienne Marcotte, Sai Rajeswar, Valentina Zantedeschi, David Vazquez, Nicolas Chapados, Christopher Pal, Perouz Taslakian | 2024-06-17T17:52:54Z | cs.CL, cs.AI | Large Language Models (LLMs) are trained on vast amounts of data, most of
which is automatically scraped from the internet. This data includes
encyclopedic documents that harbor a vast amount of general knowledge (e.g.,
Wikipedia) but also potentially overlap with benchmark datasets used for
evaluating LLMs. Consequently, evaluating models on test splits that might have
leaked into the training set is prone to misleading conclusions. To foster
sound evaluation of language models, we introduce a new test dataset named
RepLiQA, suited for question-answering and topic retrieval tasks. RepLiQA is a
collection of five splits of test sets, four of which have not been released to
the internet or exposed to LLM APIs prior to this publication. Each sample in
RepLiQA comprises (1) a reference document crafted by a human annotator and
depicting an imaginary scenario (e.g., a news article) absent from the
internet; (2) a question about the document's topic; (3) a ground-truth answer
derived directly from the information in the document; and (4) the paragraph
extracted from the reference document containing the answer. As such, accurate
answers can only be generated if a model can find relevant content within the
provided document. We run a large-scale benchmark comprising several
state-of-the-art LLMs to uncover differences in performance across models of
various types and sizes in a context-conditional language modeling setting.
Released splits of RepLiQA can be found here:
https://huggingface.co/datasets/ServiceNow/repliqa. | cs.CL | RepLiQA: A Question-Answering Dataset for Benchmarking LLMs on Unseen Reference Content |
2024-06-17T17:49:00Z | http://arxiv.org/pdf/2406.11944v1 | http://arxiv.org/abs/2406.11944v1 | Jacob Dunefsky, Philippe Chlenski, Neel Nanda | 2024-06-17T17:49:00Z | cs.LG, cs.CL | A key goal in mechanistic interpretability is circuit analysis: finding
sparse subgraphs of models corresponding to specific behaviors or capabilities.
However, MLP sublayers make fine-grained circuit analysis on transformer-based
language models difficult. In particular, interpretable features -- such as
those found by sparse autoencoders (SAEs) -- are typically linear combinations
of extremely many neurons, each with its own nonlinearity to account for.
Circuit analysis in this setting thus either yields intractably large circuits
or fails to disentangle local and global behavior. To address this we explore
transcoders, which seek to faithfully approximate a densely activating MLP
layer with a wider, sparsely-activating MLP layer. We successfully train
transcoders on language models with 120M, 410M, and 1.4B parameters, and find
them to perform at least on par with SAEs in terms of sparsity, faithfulness,
and human-interpretability. We then introduce a novel method for using
transcoders to perform weights-based circuit analysis through MLP sublayers.
The resulting circuits neatly factorize into input-dependent and
input-invariant terms. Finally, we apply transcoders to reverse-engineer
unknown circuits in the model, and we obtain novel insights regarding the
greater-than circuit in GPT2-small. Our results suggest that transcoders can
prove effective in decomposing model computations involving MLPs into
interpretable circuits. Code is available at
https://github.com/jacobdunefsky/transcoder_circuits. | cs.LG | Transcoders Find Interpretable LLM Feature Circuits |
2024-06-17T17:48:13Z | http://arxiv.org/pdf/2406.11801v1 | http://arxiv.org/abs/2406.11801v1 | Rima Hazra, Sayan Layek, Somnath Banerjee, Soujanya Poria | 2024-06-17T17:48:13Z | cs.CL | Ensuring the safe alignment of large language models (LLMs) with human values
is critical as they become integral to applications like translation and
question answering. Current alignment methods struggle with dynamic user
intentions and complex objectives, making models vulnerable to generating
harmful content. We propose Safety Arithmetic, a training-free framework
enhancing LLM safety across different scenarios: Base models, Supervised
fine-tuned models (SFT), and Edited models. Safety Arithmetic involves Harm
Direction Removal to avoid harmful content and Safety Alignment to promote safe
responses. Additionally, we present NoIntentEdit, a dataset highlighting edit
instances that could compromise model safety if used unintentionally. Our
experiments show that Safety Arithmetic significantly improves safety measures,
reduces over-safety, and maintains model utility, outperforming existing
methods in ensuring safe content generation. | cs.CL | Safety Arithmetic: A Framework for Test-time Safety Alignment of Language Models by Steering Parameters and Activations |
2024-06-17T17:42:57Z | http://arxiv.org/pdf/2406.11794v3 | http://arxiv.org/abs/2406.11794v3 | Jeffrey Li, Alex Fang, Georgios Smyrnis, Maor Ivgi, Matt Jordan, Samir Gadre, Hritik Bansal, Etash Guha, Sedrick Keh, Kushal Arora, Saurabh Garg, Rui Xin, Niklas Muennighoff, Reinhard Heckel, Jean Mercat, Mayee Chen, Suchin Gururangan, Mitchell Wortsman, Alon Albalak, Yonatan Bitton, Marianna Nezhurina, Amro Abbas, Cheng-Yu Hsieh, Dhruba Ghosh, Josh Gardner, Maciej Kilian, Hanlin Zhang, Rulin Shao, Sarah Pratt, Sunny Sanyal, Gabriel Ilharco, Giannis Daras, Kalyani Marathe, Aaron Gokaslan, Jieyu Zhang, Khyathi Chandu, Thao Nguyen, Igor Vasiljevic, Sham Kakade, Shuran Song, Sujay Sanghavi, Fartash Faghri, Sewoong Oh, Luke Zettlemoyer, Kyle Lo, Alaaeldin El-Nouby, Hadi Pouransari, Alexander Toshev, Stephanie Wang, Dirk Groeneveld, Luca Soldaini, Pang Wei Koh, Jenia Jitsev, Thomas Kollar, Alexandros G. Dimakis, Yair Carmon, Achal Dave, Ludwig Schmidt, Vaishaal Shankar | 2024-06-20T17:43:05Z | cs.LG, cs.CL | We introduce DataComp for Language Models (DCLM), a testbed for controlled
dataset experiments with the goal of improving language models. As part of
DCLM, we provide a standardized corpus of 240T tokens extracted from Common
Crawl, effective pretraining recipes based on the OpenLM framework, and a broad
suite of 53 downstream evaluations. Participants in the DCLM benchmark can
experiment with data curation strategies such as deduplication, filtering, and
data mixing at model scales ranging from 412M to 7B parameters. As a baseline
for DCLM, we conduct extensive experiments and find that model-based filtering
is key to assembling a high-quality training set. The resulting dataset,
DCLM-Baseline enables training a 7B parameter language model from scratch to
64% 5-shot accuracy on MMLU with 2.6T training tokens. Compared to MAP-Neo, the
previous state-of-the-art in open-data language models, DCLM-Baseline
represents a 6.6 percentage point improvement on MMLU while being trained with
40% less compute. Our baseline model is also comparable to Mistral-7B-v0.3 and
Llama 3 8B on MMLU (63% & 66%), and performs similarly on an average of 53
natural language understanding tasks while being trained with 6.6x less compute
than Llama 3 8B. Our results highlight the importance of dataset design for
training language models and offer a starting point for further research on
data curation. | cs.LG | DataComp-LM: In search of the next generation of training sets for language models |
2024-06-17T17:39:10Z | http://arxiv.org/pdf/2406.11785v1 | http://arxiv.org/abs/2406.11785v1 | Ronny Luss, Erik Miehling, Amit Dhurandhar | 2024-06-17T17:39:10Z | cs.CL, cs.AI, cs.LG | The advent of black-box deep neural network classification models has sparked
the need to explain their decisions. However, in the case of generative AI such
as large language models (LLMs), there is no class prediction to explain.
Rather, one can ask why an LLM output a particular response to a given prompt.
In this paper, we answer this question by proposing, to the best of our
knowledge, the first contrastive explanation methods requiring simply
black-box/query access. Our explanations suggest that an LLM outputs a reply to
a given prompt because if the prompt was slightly modified, the LLM would have
given a different response that is either less preferable or contradicts the
original response. The key insight is that contrastive explanations simply
require a distance function that has meaning to the user and not necessarily a
real valued representation of a specific response (viz. class label). We offer
two algorithms for finding contrastive explanations: i) A myopic algorithm,
which although effective in creating contrasts, requires many model calls and
ii) A budgeted algorithm, our main algorithmic contribution, which
intelligently creates contrasts adhering to a query budget, necessary for
longer contexts. We show the efficacy of these methods on diverse natural
language tasks such as open-text generation, automated red teaming, and
explaining conversational degradation. | cs.CL | CELL your Model: Contrastive Explanation Methods for Large Language Models |
2024-06-17T17:38:43Z | http://arxiv.org/pdf/2406.11784v1 | http://arxiv.org/abs/2406.11784v1 | Peter Baile Chen, Yi Zhang, Chunwei Liu, Sejal Gupta, Yoon Kim, Michael Cafarella | 2024-06-17T17:38:43Z | cs.CL, cs.AI | The same real-life questions posed to different individuals may lead to
different answers based on their unique situations. For instance, whether a
student is eligible for a scholarship depends on eligibility conditions, such
as major or degree required. ConditionalQA was proposed to evaluate models'
capability of reading a document and answering eligibility questions,
considering unmentioned conditions. However, it is limited to questions on
single documents, neglecting harder cases that may require cross-document
reasoning and optimization, for example, "What is the maximum number of
scholarships attainable?" Such questions over multiple documents are not only
more challenging due to more context having to understand, but also because the
model has to (1) explore all possible combinations of unmentioned conditions
and (2) understand the relationship between conditions across documents, to
reason about the optimal outcome. To evaluate models' capability of answering
such questions, we propose a new dataset MDCR, which can reflect real-world
challenges and serve as a new test bed for complex conditional reasoning that
requires optimization. We evaluate this dataset using the most recent LLMs and
demonstrate their limitations in solving this task. We believe this dataset
will facilitate future research in answering optimization questions with
unknown conditions. | cs.CL | MDCR: A Dataset for Multi-Document Conditional Reasoning |
2024-06-17T17:35:52Z | http://arxiv.org/pdf/2406.11780v1 | http://arxiv.org/abs/2406.11780v1 | Swanand Ravindra Kadhe, Farhan Ahmed, Dennis Wei, Nathalie Baracaldo, Inkit Padhi | 2024-06-17T17:35:52Z | cs.LG, cs.AI, cs.CL | Large language models (LLMs) have shown to pose social and ethical risks such
as generating toxic language or facilitating malicious use of hazardous
knowledge. Machine unlearning is a promising approach to improve LLM safety by
directly removing harmful behaviors and knowledge. In this paper, we propose
"SPlit, UNlearn, MerGE" (SPUNGE), a framework that can be used with any
unlearning method to amplify its effectiveness. SPUNGE leverages data
attributes during unlearning by splitting unlearning data into subsets based on
specific attribute values, unlearning each subset separately, and merging the
unlearned models. We empirically demonstrate that SPUNGE significantly improves
the performance of two recent unlearning methods on state-of-the-art LLMs while
maintaining their general capabilities on standard academic benchmarks. | cs.LG | Split, Unlearn, Merge: Leveraging Data Attributes for More Effective Unlearning in LLMs |
2024-06-17T17:33:09Z | http://arxiv.org/pdf/2406.11776v1 | http://arxiv.org/abs/2406.11776v1 | Yunxuan Li, Yibing Du, Jiageng Zhang, Le Hou, Peter Grabowski, Yeqing Li, Eugene Ie | 2024-06-17T17:33:09Z | cs.CL | Multi-agent debate has proven effective in improving large language models
quality for reasoning and factuality tasks. While various role-playing
strategies in multi-agent debates have been explored, in terms of the
communication among agents, existing approaches adopt a brute force algorithm
-- each agent can communicate with all other agents. In this paper, we
systematically investigate the effect of communication connectivity in
multi-agent systems. Our experiments on GPT and Mistral models reveal that
multi-agent debates leveraging sparse communication topology can achieve
comparable or superior performance while significantly reducing computational
costs. Furthermore, we extend the multi-agent debate framework to multimodal
reasoning and alignment labeling tasks, showcasing its broad applicability and
effectiveness. Our findings underscore the importance of communication
connectivity on enhancing the efficiency and effectiveness of the "society of
minds" approach. | cs.CL | Improving Multi-Agent Debate with Sparse Communication Topology |
2024-06-17T17:31:01Z | http://arxiv.org/pdf/2406.11768v1 | http://arxiv.org/abs/2406.11768v1 | Sreyan Ghosh, Sonal Kumar, Ashish Seth, Chandra Kiran Reddy Evuru, Utkarsh Tyagi, S Sakshi, Oriol Nieto, Ramani Duraiswami, Dinesh Manocha | 2024-06-17T17:31:01Z | cs.SD, cs.AI, cs.CL, eess.AS | Perceiving and understanding non-speech sounds and non-verbal speech is
essential to making decisions that help us interact with our surroundings. In
this paper, we propose GAMA, a novel General-purpose Large Audio-Language Model
(LALM) with Advanced Audio Understanding and Complex Reasoning Abilities. We
build GAMA by integrating an LLM with multiple types of audio representations,
including features from a custom Audio Q-Former, a multi-layer aggregator that
aggregates features from multiple layers of an audio encoder. We fine-tune GAMA
on a large-scale audio-language dataset, which augments it with audio
understanding capabilities. Next, we propose CompA-R (Instruction-Tuning for
Complex Audio Reasoning), a synthetically generated instruction-tuning (IT)
dataset with instructions that require the model to perform complex reasoning
on the input audio. We instruction-tune GAMA with CompA-R to endow it with
complex reasoning abilities, where we further add a soft prompt as input with
high-level semantic evidence by leveraging event tags of the input audio.
Finally, we also propose CompA-R-test, a human-labeled evaluation dataset for
evaluating the capabilities of LALMs on open-ended audio question-answering
that requires complex reasoning. Through automated and expert human
evaluations, we show that GAMA outperforms all other LALMs in literature on
diverse audio understanding tasks by margins of 1%-84%. Further, GAMA IT-ed on
CompA-R proves to be superior in its complex reasoning and instruction
following capabilities. | cs.SD | GAMA: A Large Audio-Language Model with Advanced Audio Understanding and Complex Reasoning Abilities |
2024-06-17T17:26:10Z | http://arxiv.org/pdf/2406.11939v1 | http://arxiv.org/abs/2406.11939v1 | Tianle Li, Wei-Lin Chiang, Evan Frick, Lisa Dunlap, Tianhao Wu, Banghua Zhu, Joseph E. Gonzalez, Ion Stoica | 2024-06-17T17:26:10Z | cs.LG, cs.AI, cs.CL | The rapid evolution of language models has necessitated the development of
more challenging benchmarks. Current static benchmarks often struggle to
consistently distinguish between the capabilities of different models and fail
to align with real-world user preferences. On the other hand, live
crowd-sourced platforms like the Chatbot Arena collect a wide range of natural
prompts and user feedback. However, these prompts vary in sophistication and
the feedback cannot be applied offline to new models. In order to ensure that
benchmarks keep up with the pace of LLM development, we address how one can
evaluate benchmarks on their ability to confidently separate models and their
alignment with human preference. Under these principles, we developed
BenchBuilder, a living benchmark that filters high-quality prompts from live
data sources to enable offline evaluation on fresh, challenging prompts.
BenchBuilder identifies seven indicators of a high-quality prompt, such as the
requirement for domain knowledge, and utilizes an LLM annotator to select a
high-quality subset of prompts from various topic clusters. The LLM evaluation
process employs an LLM judge to ensure a fully automated, high-quality, and
constantly updating benchmark. We apply BenchBuilder on prompts from the
Chatbot Arena to create Arena-Hard-Auto v0.1: 500 challenging user prompts from
a wide range of tasks. Arena-Hard-Auto v0.1 offers 3x tighter confidence
intervals than MT-Bench and achieves a state-of-the-art 89.1% agreement with
human preference rankings, all at a cost of only $25 and without human
labelers. The BenchBuilder pipeline enhances evaluation benchmarks and provides
a valuable tool for developers, enabling them to extract high-quality
benchmarks from extensive data with minimal effort. | cs.LG | From Crowdsourced Data to High-Quality Benchmarks: Arena-Hard and BenchBuilder Pipeline |
2024-06-17T17:16:45Z | http://arxiv.org/pdf/2406.11757v1 | http://arxiv.org/abs/2406.11757v1 | Laura Weidinger, John Mellor, Bernat Guillen Pegueroles, Nahema Marchal, Ravin Kumar, Kristian Lum, Canfer Akbulut, Mark Diaz, Stevie Bergman, Mikel Rodriguez, Verena Rieser, William Isaac | 2024-06-17T17:16:45Z | cs.AI, cs.CL, cs.CY, cs.HC | This research introduces STAR, a sociotechnical framework that improves on
current best practices for red teaming safety of large language models. STAR
makes two key contributions: it enhances steerability by generating
parameterised instructions for human red teamers, leading to improved coverage
of the risk surface. Parameterised instructions also provide more detailed
insights into model failures at no increased cost. Second, STAR improves signal
quality by matching demographics to assess harms for specific groups, resulting
in more sensitive annotations. STAR further employs a novel step of arbitration
to leverage diverse viewpoints and improve label reliability, treating
disagreement not as noise but as a valuable contribution to signal quality. | cs.AI | STAR: SocioTechnical Approach to Red Teaming Language Models |
2024-06-17T17:13:08Z | http://arxiv.org/pdf/2406.11753v1 | http://arxiv.org/abs/2406.11753v1 | Jian Gu, Aldeida Aleti, Chunyang Chen, Hongyu Zhang | 2024-06-17T17:13:08Z | cs.CL, cs.LG | Finetuning language models (LMs) is crucial for adapting the models to
downstream data and tasks. However, full finetuning is usually costly. Existing
work, such as parameter-efficient finetuning (PEFT), often focuses on
\textit{how to finetune} but neglects the issue of \textit{where to finetune}.
As a pioneering work on answering where to finetune (at the layer level), we
conduct a semantic analysis of the LM inference process. We first propose a
virtual transition of the latent representation and then trace its factual
transition. Based on the deviation in transitions, we estimate the gain of
finetuning each model layer, and further, narrow down the scope for finetuning.
We perform extensive experiments across well-known LMs and datasets. The
results show that our approach is effective and efficient, and outperforms the
existing baselines. Our approach is orthogonal to existing efficient
techniques, such as PEFT methods, offering practical values on LM finetuning. | cs.CL | A Semantic-based Layer Freezing Approach to Efficient Fine-Tuning of Language Models |
2024-06-17T17:02:34Z | http://arxiv.org/pdf/2406.11745v1 | http://arxiv.org/abs/2406.11745v1 | Wenjia Zhang, Lin Gui, Rob Procter, Yulan He | 2024-06-17T17:02:34Z | cs.IR, cs.CL | To seek reliable information sources for news events, we introduce a novel
task of expert recommendation, which aims to identify trustworthy sources based
on their previously quoted statements. To achieve this, we built a novel
dataset, called NewsQuote, consisting of 23,571 quote-speaker pairs sourced
from a collection of news articles. We formulate the recommendation task as the
retrieval of experts based on their likelihood of being associated with a given
query. We also propose a multi-layer ranking framework employing Large Language
Models to improve the recommendation performance. Our results show that
employing an in-context learning based LLM ranker and a multi-layer
ranking-based filter significantly improve both the predictive quality and
behavioural quality of the recommender system. | cs.IR | Multi-Layer Ranking with Large Language Models for News Source Recommendation |
2024-06-17T16:52:56Z | http://arxiv.org/pdf/2406.11736v1 | http://arxiv.org/abs/2406.11736v1 | Fangzhi Xu, Qiushi Sun, Kanzhi Cheng, Jun Liu, Yu Qiao, Zhiyong Wu | 2024-06-17T16:52:56Z | cs.CL, cs.AI | One of the primary driving forces contributing to the superior performance of
Large Language Models (LLMs) is the extensive availability of human-annotated
natural language data, which is used for alignment fine-tuning. This inspired
researchers to investigate self-training methods to mitigate the extensive
reliance on human annotations. However, the current success of self-training
has been primarily observed in natural language scenarios, rather than in the
increasingly important neural-symbolic scenarios. To this end, we propose an
environment-guided neural-symbolic self-training framework named ENVISIONS. It
aims to overcome two main challenges: (1) the scarcity of symbolic data, and
(2) the limited proficiency of LLMs in processing symbolic language. Extensive
evaluations conducted on three distinct domains demonstrate the effectiveness
of our approach. Additionally, we have conducted a comprehensive analysis to
uncover the factors contributing to ENVISIONS's success, thereby offering
valuable insights for future research in this area. Code will be available at
\url{https://github.com/xufangzhi/ENVISIONS}. | cs.CL | Interactive Evolution: A Neural-Symbolic Self-Training Framework For Large Language Models |
2024-06-17T16:46:10Z | http://arxiv.org/pdf/2406.11727v2 | http://arxiv.org/abs/2406.11727v2 | Sewade Ogun, Abraham T. Owodunni, Tobi Olatunji, Eniola Alese, Babatunde Oladimeji, Tejumade Afonja, Kayode Olaleye, Naome A. Etori, Tosin Adewumi | 2024-06-27T08:52:54Z | eess.AS, cs.CL | Recent advances in speech synthesis have enabled many useful applications
like audio directions in Google Maps, screen readers, and automated content
generation on platforms like TikTok. However, these systems are mostly
dominated by voices sourced from data-rich geographies with personas
representative of their source data. Although 3000 of the world's languages are
domiciled in Africa, African voices and personas are under-represented in these
systems. As speech synthesis becomes increasingly democratized, it is desirable
to increase the representation of African English accents. We present Afro-TTS,
the first pan-African accented English speech synthesis system able to generate
speech in 86 African accents, with 1000 personas representing the rich
phonological diversity across the continent for downstream application in
Education, Public Health, and Automated Content Creation. Speaker interpolation
retains naturalness and accentedness, enabling the creation of new voices. | eess.AS | 1000 African Voices: Advancing inclusive multi-speaker multi-accent speech synthesis |
2024-06-17T16:40:21Z | http://arxiv.org/pdf/2406.11721v1 | http://arxiv.org/abs/2406.11721v1 | Bingxiang He, Ning Ding, Cheng Qian, Jia Deng, Ganqu Cui, Lifan Yuan, Huan-ang Gao, Huimin Chen, Zhiyuan Liu, Maosong Sun | 2024-06-17T16:40:21Z | cs.CL, cs.AI, cs.LG | Understanding alignment techniques begins with comprehending zero-shot
generalization brought by instruction tuning, but little of the mechanism has
been understood. Existing work has largely been confined to the task level,
without considering that tasks are artificially defined and, to LLMs, merely
consist of tokens and representations. This line of research has been limited
to examining transfer between tasks from a task-pair perspective, with few
studies focusing on understanding zero-shot generalization from the perspective
of the data itself. To bridge this gap, we first demonstrate through multiple
metrics that zero-shot generalization during instruction tuning happens very
early. Next, we investigate the facilitation of zero-shot generalization from
both data similarity and granularity perspectives, confirming that encountering
highly similar and fine-grained training data earlier during instruction
tuning, without the constraints of defined "tasks", enables better
generalization. Finally, we propose a more grounded training data arrangement
method, Test-centric Multi-turn Arrangement, and show its effectiveness in
promoting continual learning and further loss reduction. For the first time, we
show that zero-shot generalization during instruction tuning is a form of
similarity-based generalization between training and test data at the instance
level. We hope our analysis will advance the understanding of zero-shot
generalization during instruction tuning and contribute to the development of
more aligned LLMs. Our code is released at
https://github.com/HBX-hbx/dynamics_of_zero-shot_generalization. | cs.CL | Zero-Shot Generalization during Instruction Tuning: Insights from Similarity and Granularity |
2024-06-17T16:36:12Z | http://arxiv.org/pdf/2406.11717v1 | http://arxiv.org/abs/2406.11717v1 | Andy Arditi, Oscar Obeso, Aaquib Syed, Daniel Paleka, Nina Rimsky, Wes Gurnee, Neel Nanda | 2024-06-17T16:36:12Z | cs.LG, cs.AI, cs.CL | Conversational large language models are fine-tuned for both
instruction-following and safety, resulting in models that obey benign requests
but refuse harmful ones. While this refusal behavior is widespread across chat
models, its underlying mechanisms remain poorly understood. In this work, we
show that refusal is mediated by a one-dimensional subspace, across 13 popular
open-source chat models up to 72B parameters in size. Specifically, for each
model, we find a single direction such that erasing this direction from the
model's residual stream activations prevents it from refusing harmful
instructions, while adding this direction elicits refusal on even harmless
instructions. Leveraging this insight, we propose a novel white-box jailbreak
method that surgically disables refusal with minimal effect on other
capabilities. Finally, we mechanistically analyze how adversarial suffixes
suppress propagation of the refusal-mediating direction. Our findings
underscore the brittleness of current safety fine-tuning methods. More broadly,
our work showcases how an understanding of model internals can be leveraged to
develop practical methods for controlling model behavior. | cs.LG | Refusal in Language Models Is Mediated by a Single Direction |
2024-06-17T16:33:35Z | http://arxiv.org/pdf/2406.11715v1 | http://arxiv.org/abs/2406.11715v1 | Aneesh Pappu, Billy Porter, Ilia Shumailov, Jamie Hayes | 2024-06-17T16:33:35Z | cs.LG, cs.CL, cs.SE | Reinforcement learning with human feedback (RLHF) has become the dominant
method to align large models to user preferences. Unlike fine-tuning, for which
there are many studies regarding training data memorization, it is not clear
how memorization is affected by or introduced in the RLHF alignment process.
Understanding this relationship is important as real user data may be collected
and used to align large models; if user data is memorized during RLHF and later
regurgitated, this could raise privacy concerns. In this work, we analyze how
training data memorization can surface and propagate through each phase of
RLHF. We focus our study on code completion models, as code completion is one
of the most popular use cases for large language models. We find that RLHF
significantly decreases the chance that data used for reward modeling and
reinforcement learning is memorized, in comparison to aligning via directly
fine-tuning on this data, but that examples already memorized during the
fine-tuning stage of RLHF, will, in the majority of cases, remain memorized
after RLHF. | cs.LG | Measuring memorization in RLHF for code completion |
2024-06-17T16:28:21Z | http://arxiv.org/pdf/2406.11709v2 | http://arxiv.org/abs/2406.11709v2 | Priyanka Kargupta, Ishika Agarwal, Dilek Hakkani-Tur, Jiawei Han | 2024-06-25T21:17:41Z | cs.CL, cs.MA | Socratic questioning is an effective teaching strategy, encouraging critical
thinking and problem-solving. The conversational capabilities of large language
models (LLMs) show great potential for providing scalable, real-time student
guidance. However, current LLMs often give away solutions directly, making them
ineffective instructors. We tackle this issue in the code debugging domain with
TreeInstruct, an Instructor agent guided by a novel state space-based planning
algorithm. TreeInstruct asks probing questions to help students independently
identify and resolve errors. It estimates a student's conceptual and
syntactical knowledge to dynamically construct a question tree based on their
responses and current knowledge state, effectively addressing both independent
and dependent mistakes concurrently in a multi-turn interaction setting. In
addition to using an existing single-bug debugging benchmark, we construct a
more challenging multi-bug dataset of 150 coding problems, incorrect solutions,
and bug fixes -- all carefully constructed and annotated by experts. Extensive
evaluation shows TreeInstruct's state-of-the-art performance on both datasets,
proving it to be a more effective instructor than baselines. Furthermore, a
real-world case study with five students of varying skill levels further
demonstrates TreeInstruct's ability to guide students to debug their code
efficiently with minimal turns and highly Socratic questioning. | cs.CL | Instruct, Not Assist: LLM-based Multi-Turn Planning and Hierarchical Questioning for Socratic Code Debugging |
2024-06-17T16:25:55Z | http://arxiv.org/pdf/2406.11706v1 | http://arxiv.org/abs/2406.11706v1 | Jasper Xian, Saron Samuel, Faraz Khoubsirat, Ronak Pradeep, Md Arafat Sultan, Radu Florian, Salim Roukos, Avirup Sil, Christopher Potts, Omar Khattab | 2024-06-17T16:25:55Z | cs.IR, cs.CL, cs.LG | We develop a method for training small-scale (under 100M parameter) neural
information retrieval models with as few as 10 gold relevance labels. The
method depends on generating synthetic queries for documents using a language
model (LM), and the key step is that we automatically optimize the LM prompt
that is used to generate these queries based on training quality. In
experiments with the BIRCO benchmark, we find that models trained with our
method outperform RankZephyr and are competitive with RankLLama, both of which
are 7B parameter models trained on over 100K labels. These findings point to
the power of automatic prompt optimization for synthetic dataset generation. | cs.IR | Prompts as Auto-Optimized Training Hyperparameters: Training Best-in-Class IR Models from Scratch with 10 Gold Labels |
2024-06-17T16:25:04Z | http://arxiv.org/pdf/2406.11704v1 | http://arxiv.org/abs/2406.11704v1 | Nvidia, :, Bo Adler, Niket Agarwal, Ashwath Aithal, Dong H. Anh, Pallab Bhattacharya, Annika Brundyn, Jared Casper, Bryan Catanzaro, Sharon Clay, Jonathan Cohen, Sirshak Das, Ayush Dattagupta, Olivier Delalleau, Leon Derczynski, Yi Dong, Daniel Egert, Ellie Evans, Aleksander Ficek, Denys Fridman, Shaona Ghosh, Boris Ginsburg, Igor Gitman, Tomasz Grzegorzek, Robert Hero, Jining Huang, Vibhu Jawa, Joseph Jennings, Aastha Jhunjhunwala, John Kamalu, Sadaf Khan, Oleksii Kuchaiev, Patrick LeGresley, Hui Li, Jiwei Liu, Zihan Liu, Eileen Long, Ameya Sunil Mahabaleshwarkar, Somshubra Majumdar, James Maki, Miguel Martinez, Maer Rodrigues de Melo, Ivan Moshkov, Deepak Narayanan, Sean Narenthiran, Jesus Navarro, Phong Nguyen, Osvald Nitski, Vahid Noroozi, Guruprasad Nutheti, Christopher Parisien, Jupinder Parmar, Mostofa Patwary, Krzysztof Pawelec, Wei Ping, Shrimai Prabhumoye, Rajarshi Roy, Trisha Saar, Vasanth Rao Naik Sabavat, Sanjeev Satheesh, Jane Polak Scowcroft, Jason Sewall, Pavel Shamis, Gerald Shen, Mohammad Shoeybi, Dave Sizer, Misha Smelyanskiy, Felipe Soares, Makesh Narsimhan Sreedhar, Dan Su, Sandeep Subramanian, Shengyang Sun, Shubham Toshniwal, Hao Wang, Zhilin Wang, Jiaxuan You, Jiaqi Zeng, Jimmy Zhang, Jing Zhang, Vivienne Zhang, Yian Zhang, Chen Zhu | 2024-06-17T16:25:04Z | cs.CL, cs.AI, cs.LG | We release the Nemotron-4 340B model family, including Nemotron-4-340B-Base,
Nemotron-4-340B-Instruct, and Nemotron-4-340B-Reward. Our models are open
access under the NVIDIA Open Model License Agreement, a permissive model
license that allows distribution, modification, and use of the models and its
outputs. These models perform competitively to open access models on a wide
range of evaluation benchmarks, and were sized to fit on a single DGX H100 with
8 GPUs when deployed in FP8 precision. We believe that the community can
benefit from these models in various research studies and commercial
applications, especially for generating synthetic data to train smaller
language models. Notably, over 98% of data used in our model alignment process
is synthetically generated, showcasing the effectiveness of these models in
generating synthetic data. To further support open research and facilitate
model development, we are also open-sourcing the synthetic data generation
pipeline used in our model alignment process. | cs.CL | Nemotron-4 340B Technical Report |
2024-06-17T16:14:11Z | http://arxiv.org/pdf/2406.11698v1 | http://arxiv.org/abs/2406.11698v1 | Peizhong Gao, Ao Xie, Shaoguang Mao, Wenshan Wu, Yan Xia, Haipeng Mi, Furu Wei | 2024-06-17T16:14:11Z | cs.CL | We introduce Meta-Reasoning Prompting (MRP), a novel and efficient system
prompting method for large language models (LLMs) inspired by human
meta-reasoning. Traditional in-context learning-based reasoning techniques,
such as Tree-of-Thoughts, show promise but lack consistent state-of-the-art
performance across diverse tasks due to their specialized nature. MRP addresses
this limitation by guiding LLMs to dynamically select and apply different
reasoning methods based on the specific requirements of each task, optimizing
both performance and computational efficiency. With MRP, LLM reasoning operates
in two phases. Initially, the LLM identifies the most appropriate reasoning
method using task input cues and objective descriptions of available methods.
Subsequently, it applies the chosen method to complete the task. This dynamic
strategy mirrors human meta-reasoning, allowing the model to excel in a wide
range of problem domains. We evaluate the effectiveness of MRP through
comprehensive benchmarks. The results demonstrate that MRP achieves or
approaches state-of-the-art performance across diverse tasks. MRP represents a
significant advancement in enabling LLMs to identify cognitive challenges
across problems and leverage benefits across different reasoning approaches,
enhancing their ability to handle diverse and complex problem domains
efficiently. Every LLM deserves a Meta-Reasoning Prompting to unlock its full
potential and ensure adaptability in an ever-evolving landscape of challenges
and applications. | cs.CL | Meta Reasoning for Large Language Models |
2024-06-17T16:12:03Z | http://arxiv.org/pdf/2406.11695v1 | http://arxiv.org/abs/2406.11695v1 | Krista Opsahl-Ong, Michael J Ryan, Josh Purtell, David Broman, Christopher Potts, Matei Zaharia, Omar Khattab | 2024-06-17T16:12:03Z | cs.CL, cs.AI, cs.LG | Language Model Programs, i.e. sophisticated pipelines of modular language
model (LM) calls, are increasingly advancing NLP tasks, but they require
crafting prompts that are jointly effective for all modules. We study prompt
optimization for LM programs, i.e. how to update these prompts to maximize a
downstream metric without access to module-level labels or gradients. To make
this tractable, we factorize our problem into optimizing the free-form
instructions and few-shot demonstrations of every module and introduce several
strategies to craft task-grounded instructions and navigate credit assignment
across modules. Our strategies include (i) program- and data-aware techniques
for proposing effective instructions, (ii) a stochastic mini-batch evaluation
function for learning a surrogate model of our objective, and (iii) a
meta-optimization procedure in which we refine how LMs construct proposals over
time. Using these insights we develop MIPRO, a novel optimizer that outperforms
baselines on five of six diverse LM programs using a best-in-class open-source
model (Llama-3-8B), by as high as 12.9% accuracy. We will release our new
optimizers and benchmark in DSPy at https://github.com/stanfordnlp/dspy | cs.CL | Optimizing Instructions and Demonstrations for Multi-Stage Language Model Programs |
2024-06-17T16:05:32Z | http://arxiv.org/pdf/2406.11687v1 | http://arxiv.org/abs/2406.11687v1 | Yekun Chai, Yewei Fang, Qiwei Peng, Xuhong Li | 2024-06-17T16:05:32Z | cs.CL | Language models typically tokenize raw text into sequences of subword
identifiers from a predefined vocabulary, a process inherently sensitive to
typographical errors, length variations, and largely oblivious to the internal
structure of tokens-issues we term the curse of tokenization. In this study, we
delve into these drawbacks and demonstrate that large language models (LLMs)
remain susceptible to these problems. This study systematically investigates
these challenges and their impact on LLMs through three critical research
questions: (1) complex problem solving, (2) token structure probing, and (3)
resilience to typographical variation. Our findings reveal that scaling model
parameters can mitigate the issue of tokenization; however, LLMs still suffer
from biases induced by typos and other text format variations. Our experiments
show that subword regularization such as BPE-dropout can mitigate this issue.
We will release our code and data to facilitate further research. | cs.CL | Tokenization Falling Short: The Curse of Tokenization |
2024-06-17T16:01:33Z | http://arxiv.org/pdf/2406.11683v1 | http://arxiv.org/abs/2406.11683v1 | Jing Chen, Xinyu Zhu, Cheng Yang, Chufan Shi, Yadong Xi, Yuxiang Zhang, Junjie Wang, Jiashu Pu, Rongsheng Zhang, Yujiu Yang, Tian Feng | 2024-06-17T16:01:33Z | cs.CL | Generative AI has demonstrated unprecedented creativity in the field of
computer vision, yet such phenomena have not been observed in natural language
processing. In particular, large language models (LLMs) can hardly produce
written works at the level of human experts due to the extremely high
complexity of literature writing. In this paper, we present HoLLMwood, an
automated framework for unleashing the creativity of LLMs and exploring their
potential in screenwriting, which is a highly demanding task. Mimicking the
human creative process, we assign LLMs to different roles involved in the
real-world scenario. In addition to the common practice of treating LLMs as
${Writer}$, we also apply LLMs as ${Editor}$, who is responsible for providing
feedback and revision advice to ${Writer}$. Besides, to enrich the characters
and deepen the plots, we introduce a role-playing mechanism and adopt LLMs as
${Actors}$ that can communicate and interact with each other. Evaluations on
automatically generated screenplays show that HoLLMwood substantially
outperforms strong baselines in terms of coherence, relevance, interestingness
and overall quality. | cs.CL | HoLLMwood: Unleashing the Creativity of Large Language Models in Screenwriting via Role Playing |
2024-06-17T15:59:59Z | http://arxiv.org/pdf/2406.11682v1 | http://arxiv.org/abs/2406.11682v1 | Shangqing Tu, Zhuoran Pan, Wenxuan Wang, Zhexin Zhang, Yuliang Sun, Jifan Yu, Hongning Wang, Lei Hou, Juanzi Li | 2024-06-17T15:59:59Z | cs.CL, cs.AI, cs.CR | Large language models (LLMs) have been increasingly applied to various
domains, which triggers increasing concerns about LLMs' safety on specialized
domains, e.g. medicine. However, testing the domain-specific safety of LLMs is
challenging due to the lack of domain knowledge-driven attacks in existing
benchmarks. To bridge this gap, we propose a new task, knowledge-to-jailbreak,
which aims to generate jailbreaks from domain knowledge to evaluate the safety
of LLMs when applied to those domains. We collect a large-scale dataset with
12,974 knowledge-jailbreak pairs and fine-tune a large language model as
jailbreak-generator, to produce domain knowledge-specific jailbreaks.
Experiments on 13 domains and 8 target LLMs demonstrate the effectiveness of
jailbreak-generator in generating jailbreaks that are both relevant to the
given knowledge and harmful to the target LLMs. We also apply our method to an
out-of-domain knowledge base, showing that jailbreak-generator can generate
jailbreaks that are comparable in harmfulness to those crafted by human
experts. Data and code: https://github.com/THU-KEG/Knowledge-to-Jailbreak/. | cs.CL | Knowledge-to-Jailbreak: One Knowledge Point Worth One Attack |
2024-06-17T15:59:49Z | http://arxiv.org/pdf/2406.11681v1 | http://arxiv.org/abs/2406.11681v1 | Shangqing Tu, Yuanchun Wang, Jifan Yu, Yuyang Xie, Yaran Shi, Xiaozhi Wang, Jing Zhang, Lei Hou, Juanzi Li | 2024-06-17T15:59:49Z | cs.CL, cs.AI | Large language models have achieved remarkable success on general NLP tasks,
but they may fall short for domain-specific problems. Recently, various
Retrieval-Augmented Large Language Models (RALLMs) are proposed to address this
shortcoming. However, existing evaluation tools only provide a few baselines
and evaluate them on various domains without mining the depth of domain
knowledge. In this paper, we address the challenges of evaluating RALLMs by
introducing the R-Eval toolkit, a Python toolkit designed to streamline the
evaluation of different RAG workflows in conjunction with LLMs. Our toolkit,
which supports popular built-in RAG workflows and allows for the incorporation
of customized testing data on the specific domain, is designed to be
user-friendly, modular, and extensible. We conduct an evaluation of 21 RALLMs
across three task levels and two representative domains, revealing significant
variations in the effectiveness of RALLMs across different tasks and domains.
Our analysis emphasizes the importance of considering both task and domain
requirements when choosing a RAG workflow and LLM combination. We are committed
to continuously maintaining our platform at https://github.com/THU-KEG/R-Eval
to facilitate both the industry and the researchers. | cs.CL | R-Eval: A Unified Toolkit for Evaluating Domain Knowledge of Retrieval Augmented Large Language Models |
2024-06-17T15:58:22Z | http://arxiv.org/pdf/2406.11678v1 | http://arxiv.org/abs/2406.11678v1 | Yiqun Chen, Qi Liu, Yi Zhang, Weiwei Sun, Daiting Shi, Jiaxin Mao, Dawei Yin | 2024-06-17T15:58:22Z | cs.IR, cs.CL | Large Language Models (LLMs) are increasingly employed in zero-shot documents
ranking, yielding commendable results. However, several significant challenges
still persist in LLMs for ranking: (1) LLMs are constrained by limited input
length, precluding them from processing a large number of documents
simultaneously; (2) The output document sequence is influenced by the input
order of documents, resulting in inconsistent ranking outcomes; (3) Achieving a
balance between cost and ranking performance is quite challenging. To tackle
these issues, we introduce a novel documents ranking method called TourRank,
which is inspired by the tournament mechanism. This approach alleviates the
impact of LLM's limited input length through intelligent grouping, while the
tournament-like points system ensures robust ranking, mitigating the influence
of the document input sequence. We test TourRank with different LLMs on the
TREC DL datasets and the BEIR benchmark. Experimental results show that
TourRank achieves state-of-the-art performance at a reasonable cost. | cs.IR | TourRank: Utilizing Large Language Models for Documents Ranking with a Tournament-Inspired Strategy |
2024-06-17T15:55:38Z | http://arxiv.org/pdf/2406.11675v2 | http://arxiv.org/abs/2406.11675v2 | Yibin Wang, Haizhou Shi, Ligong Han, Dimitris Metaxas, Hao Wang | 2024-06-18T15:15:04Z | cs.LG, cs.AI, cs.CL, stat.ML | Large Language Models (LLMs) often suffer from overconfidence during
inference, particularly when adapted to downstream domain-specific tasks with
limited data. Previous work addresses this issue by employing approximate
Bayesian estimation after the LLMs are trained, enabling them to quantify
uncertainty. However, such post-training approaches' performance is severely
limited by the parameters learned during training. In this paper, we go beyond
post-training Bayesianization and propose Bayesian Low-Rank Adaptation by
Backpropagation (BLoB), an algorithm that continuously and jointly adjusts both
the mean and covariance of LLM parameters throughout the whole fine-tuning
process. Our empirical results verify the effectiveness of BLoB in terms of
generalization and uncertainty estimation, when evaluated on both
in-distribution and out-of-distribution data. | cs.LG | BLoB: Bayesian Low-Rank Adaptation by Backpropagation for Large Language Models |
2024-06-17T15:55:08Z | http://arxiv.org/pdf/2406.11674v1 | http://arxiv.org/abs/2406.11674v1 | Donghyeon Joo, Ramyad Hadidi, Soheil Feizi, Bahar Asgari | 2024-06-17T15:55:08Z | cs.CL | The increasing size of large language models (LLMs) challenges their usage on
resource-constrained platforms. For example, memory on modern GPUs is
insufficient to hold LLMs that are hundreds of Gigabytes in size. Offloading is
a popular method to escape this constraint by storing weights of an LLM model
to host CPU memory and SSD, then loading each weight to GPU before every use.
In our case study of offloaded inference, we found that due to the low
bandwidth between storage devices and GPU, the latency of transferring large
model weights from its offloaded location to GPU memory becomes the critical
bottleneck with actual compute taking nearly 0% of runtime. To effectively
reduce the weight transfer latency, we propose a novel sparse format that
compresses the unstructured sparse pattern of pruned LLM weights to non-zero
values with high compression ratio and low decompression overhead. Endor
achieves this by expressing the positions of non-zero elements with a bitmap.
Compared to offloaded inference using the popular Huggingface Accelerate,
applying Endor accelerates OPT-66B by 1.70x and Llama2-70B by 1.78x. When
direct weight transfer from SSD to GPU is leveraged, Endor achieves 2.25x
speedup on OPT-66B and 2.37x speedup on Llama2-70B. | cs.CL | Endor: Hardware-Friendly Sparse Format for Offloaded LLM Inference |
2024-06-17T15:51:46Z | http://arxiv.org/pdf/2406.11670v1 | http://arxiv.org/abs/2406.11670v1 | Thorsten Pröhl, Erik Putzier, Rüdiger Zarnekow | 2024-06-17T15:51:46Z | cs.CL, cs.AI | This article gives an overview of the field of LLM text recognition.
Different approaches and implemented detectors for the recognition of
LLM-generated text are presented. In addition to discussing the
implementations, the article focuses on benchmarking the detectors. Although
there are numerous software products for the recognition of LLM-generated text,
with a focus on ChatGPT-like LLMs, the quality of the recognition (recognition
rate) is not clear. Furthermore, while it can be seen that scientific
contributions presenting their novel approaches strive for some kind of
comparison with other approaches, the construction and independence of the
evaluation dataset is often not comprehensible. As a result, discrepancies in
the performance evaluation of LLM detectors are often visible due to the
different benchmarking datasets. This article describes the creation of an
evaluation dataset and uses this dataset to investigate the different
detectors. The selected detectors are benchmarked against each other. | cs.CL | Benchmarking of LLM Detection: Comparing Two Competing Approaches |
2024-06-17T15:51:01Z | http://arxiv.org/pdf/2406.11668v1 | http://arxiv.org/abs/2406.11668v1 | Lingrui Mei, Shenghua Liu, Yiwei Wang, Baolong Bi, Jiayi Mao, Xueqi Cheng | 2024-06-17T15:51:01Z | cs.CL | "Jailbreak" is a major safety concern of Large Language Models (LLMs), which
occurs when malicious prompts lead LLMs to produce harmful outputs, raising
issues about the reliability and safety of LLMs. Therefore, an effective
evaluation of jailbreaks is very crucial to develop its mitigation strategies.
However, our research reveals that many jailbreaks identified by current
evaluations may actually be hallucinations-erroneous outputs that are mistaken
for genuine safety breaches. This finding suggests that some perceived
vulnerabilities might not represent actual threats, indicating a need for more
precise red teaming benchmarks. To address this problem, we propose the
$\textbf{B}$enchmark for reli$\textbf{AB}$ilit$\textbf{Y}$ and
jail$\textbf{B}$reak ha$\textbf{L}$l$\textbf{U}$cination $\textbf{E}$valuation
(BabyBLUE). BabyBLUE introduces a specialized validation framework including
various evaluators to enhance existing jailbreak benchmarks, ensuring outputs
are useful malicious instructions. Additionally, BabyBLUE presents a new
dataset as an augmentation to the existing red teaming benchmarks, specifically
addressing hallucinations in jailbreaks, aiming to evaluate the true potential
of jailbroken LLM outputs to cause harm to human society. | cs.CL | "Not Aligned" is Not "Malicious": Being Careful about Hallucinations of Large Language Models' Jailbreak |
2024-06-17T15:49:51Z | http://arxiv.org/pdf/2406.11665v1 | http://arxiv.org/abs/2406.11665v1 | Amith Ananthram, Elias Stengel-Eskin, Carl Vondrick, Mohit Bansal, Kathleen McKeown | 2024-06-17T15:49:51Z | cs.CL, cs.AI, cs.CV | Vision-language models (VLMs) can respond to queries about images in many
languages. However, beyond language, culture affects how we see things. For
example, individuals from Western cultures focus more on the central figure in
an image while individuals from Eastern cultures attend more to scene context.
In this work, we present a novel investigation that demonstrates and localizes
VLMs' Western bias in image understanding. We evaluate large VLMs across
subjective and objective visual tasks with culturally diverse images and
annotations. We find that VLMs perform better on the Western subset than the
Eastern subset of each task. Controlled experimentation tracing the source of
this bias highlights the importance of a diverse language mix in text-only
pre-training for building equitable VLMs, even when inference is performed in
English. Moreover, while prompting in the language of a target culture can lead
to reductions in bias, it is not a substitute for building AI more
representative of the world's languages. | cs.CL | See It from My Perspective: Diagnosing the Western Cultural Bias of Large Vision-Language Models in Image Understanding |
2024-06-17T15:43:45Z | http://arxiv.org/pdf/2406.11661v2 | http://arxiv.org/abs/2406.11661v2 | Sagnik Mukherjee, Muhammad Farid Adilazuarda, Sunayana Sitaram, Kalika Bali, Alham Fikri Aji, Monojit Choudhury | 2024-06-20T08:25:08Z | cs.CL | Socio-demographic prompting is a commonly employed approach to study cultural
biases in LLMs as well as for aligning models to certain cultures. In this
paper, we systematically probe four LLMs (Llama 3, Mistral v0.2, GPT-3.5 Turbo
and GPT-4) with prompts that are conditioned on culturally sensitive and
non-sensitive cues, on datasets that are supposed to be culturally sensitive
(EtiCor and CALI) or neutral (MMLU and ETHICS). We observe that all models
except GPT-4 show significant variations in their responses on both kinds of
datasets for both kinds of prompts, casting doubt on the robustness of the
culturally-conditioned prompting as a method for eliciting cultural bias in
models or as an alignment strategy. The work also calls rethinking the control
experiment design to tease apart the cultural conditioning of responses from
"placebo effect", i.e., random perturbations of model responses due to
arbitrary tokens in the prompt. | cs.CL | Cultural Conditioning or Placebo? On the Effectiveness of Socio-Demographic Prompting |
2024-06-17T15:41:30Z | http://arxiv.org/pdf/2406.11657v1 | http://arxiv.org/abs/2406.11657v1 | Yijiang River Dong, Tiancheng Hu, Nigel Collier | 2024-06-17T15:41:30Z | cs.CL, cs.CY | Ensuring that large language models (LLMs) reflect diverse user values and
preferences is crucial as their user bases expand globally. It is therefore
encouraging to see the growing interest in LLM personalization within the
research community. However, current works often rely on the LLM-as-a-Judge
approach for evaluation without thoroughly examining its validity. In this
paper, we investigate the reliability of LLM-as-a-Personalized-Judge, asking
LLMs to judge user preferences based on personas. Our findings suggest that
directly applying LLM-as-a-Personalized-Judge is less reliable than previously
assumed, showing low and inconsistent agreement with human ground truth. The
personas typically used are often overly simplistic, resulting in low
predictive power. To address these issues, we introduce verbal uncertainty
estimation into the LLM-as-a-Personalized-Judge pipeline, allowing the model to
express low confidence on uncertain judgments. This adjustment leads to much
higher agreement (above 80%) on high-certainty samples for binary tasks.
Through human evaluation, we find that the LLM-as-a-Personalized-Judge achieves
comparable performance to third-party humans evaluation and even surpasses
human performance on high-certainty samples. Our work indicates that
certainty-enhanced LLM-as-a-Personalized-Judge offers a promising direction for
developing more reliable and scalable methods for evaluating LLM
personalization. | cs.CL | Can LLM be a Personalized Judge? |
2024-06-17T15:36:14Z | http://arxiv.org/pdf/2406.11654v1 | http://arxiv.org/abs/2406.11654v1 | Vernon Toh Yan Han, Rishabh Bhardwaj, Soujanya Poria | 2024-06-17T15:36:14Z | cs.CL | We propose Ruby Teaming, a method that improves on Rainbow Teaming by
including a memory cache as its third dimension. The memory dimension provides
cues to the mutator to yield better-quality prompts, both in terms of attack
success rate (ASR) and quality diversity. The prompt archive generated by Ruby
Teaming has an ASR of 74%, which is 20% higher than the baseline. In terms of
quality diversity, Ruby Teaming outperforms Rainbow Teaming by 6% and 3% on
Shannon's Evenness Index (SEI) and Simpson's Diversity Index (SDI),
respectively. | cs.CL | Ruby Teaming: Improving Quality Diversity Search with Memory for Automated Red Teaming |
2024-06-17T15:32:17Z | http://arxiv.org/pdf/2406.11651v1 | http://arxiv.org/abs/2406.11651v1 | Ming Gu, Yan Yang | 2024-06-17T15:32:17Z | cs.CL | Dialogue state tracking (DST) is evaluated by exact matching methods, which
rely on large amounts of labeled data and ignore semantic consistency, leading
to over-evaluation. Currently, leveraging large language models (LLM) in
evaluating natural language processing tasks has achieved promising results.
However, using LLM for DST evaluation is still under explored. In this paper,
we propose a two-dimensional zero-shot evaluation method for DST using GPT-4,
which divides the evaluation into two dimensions: accuracy and completeness.
Furthermore, we also design two manual reasoning paths in prompting to further
improve the accuracy of evaluation. Experimental results show that our method
achieves better performance compared to the baselines, and is consistent with
traditional exact matching based methods. | cs.CL | A Two-dimensional Zero-shot Dialogue State Tracking Evaluation Method using GPT-4 |
2024-06-17T15:21:35Z | http://arxiv.org/pdf/2407.00066v1 | http://arxiv.org/abs/2407.00066v1 | Rickard Brüel-Gabrielsson, Jiacheng Zhu, Onkar Bhardwaj, Leshem Choshen, Kristjan Greenewald, Mikhail Yurochkin, Justin Solomon | 2024-06-17T15:21:35Z | cs.DC, cs.AI, cs.CL, cs.LG | Fine-tuning large language models (LLMs) with low-rank adapters (LoRAs) has
become common practice, often yielding numerous copies of the same LLM
differing only in their LoRA updates. This paradigm presents challenges for
systems that serve real-time responses to queries that each involve a different
LoRA. Prior works optimize the design of such systems but still require
continuous loading and offloading of LoRAs, as it is infeasible to store
thousands of LoRAs in GPU memory. To mitigate this issue, we investigate the
efficacy of compression when serving LoRA adapters. We consider compressing
adapters individually via SVD and propose a method for joint compression of
LoRAs into a shared basis paired with LoRA-specific scaling matrices. Our
experiments with up to 500 LoRAs demonstrate that compressed LoRAs preserve
performance while offering major throughput gains in realistic serving
scenarios with over a thousand LoRAs, maintaining 75% of the throughput of
serving a single LoRA. | cs.DC | Compress then Serve: Serving Thousands of LoRA Adapters with Little Overhead |
2024-06-17T15:14:10Z | http://arxiv.org/pdf/2406.11634v1 | http://arxiv.org/abs/2406.11634v1 | Kyle Moore, Jesse Roberts, Thao Pham, Oseremhen Ewaleifoh, Doug Fisher | 2024-06-17T15:14:10Z | cs.CL, cs.AI | Cloze testing is a common method for measuring the behavior of large language
models on a number of benchmark tasks. Using the MMLU dataset, we show that the
base-rate probability (BRP) differences across answer tokens are significant
and affect task performance ie. guess A if uncertain. We find that
counterfactual prompting does sufficiently mitigate the BRP effect. The BRP
effect is found to have a similar effect to test taking strategies employed by
humans leading to the conflation of task performance and test-taking ability.
We propose the Nvr-X-MMLU task, a variation of MMLU, which helps to
disambiguate test-taking ability from task performance and reports the latter. | cs.CL | The Base-Rate Effect on LLM Benchmark Performance: Disambiguating Test-Taking Strategies from Benchmark Performance |
2024-06-17T15:13:52Z | http://arxiv.org/pdf/2406.11632v1 | http://arxiv.org/abs/2406.11632v1 | Boxuan Lyu, Hidetaka Kamigaito, Kotaro Funakoshi, Manabu Okumura | 2024-06-17T15:13:52Z | cs.CL, cs.AI | Maximum a posteriori decoding, a commonly used method for neural machine
translation (NMT), aims to maximize the estimated posterior probability.
However, high estimated probability does not always lead to high translation
quality. Minimum Bayes Risk (MBR) decoding offers an alternative by seeking
hypotheses with the highest expected utility.
In this work, we show that Quality Estimation (QE) reranking, which uses a QE
model as a reranker, can be viewed as a variant of MBR. Inspired by this, we
propose source-based MBR (sMBR) decoding, a novel approach that utilizes
synthetic sources generated by backward translation as ``support hypotheses''
and a reference-free quality estimation metric as the utility function, marking
the first work to solely use sources in MBR decoding. Experiments show that
sMBR significantly outperforms QE reranking and is competitive with standard
MBR decoding. Furthermore, sMBR calls the utility function fewer times compared
to MBR. Our findings suggest that sMBR is a promising approach for high-quality
NMT decoding. | cs.CL | Unveiling the Power of Source: Source-based Minimum Bayes Risk Decoding for Neural Machine Translation |
2024-06-17T15:11:58Z | http://arxiv.org/pdf/2406.11629v3 | http://arxiv.org/abs/2406.11629v3 | Mingyang Song, Mao Zheng, Xuan Luo | 2024-06-30T13:31:24Z | cs.CL | Leveraging Large Language Models (LLMs) as judges for judging the performance
of LLMs has recently garnered attention. However, this type of approach is
affected by the potential biases in LLMs, raising concerns about the
reliability of the evaluation results. To mitigate this issue, we propose and
study two versions of many-shot in-context prompts, which rely on two existing
settings of many-shot ICL for helping GPT-4o-as-a-Judge in single answer
grading to mitigate the potential biases in LLMs, Reinforced ICL and
Unsupervised ICL. Concretely, the former utilizes in-context examples with
model-generated rationales, and the latter without. Based on the designed
prompts, we investigate the impact of scaling the number of in-context examples
on the consistency and quality of the judgment results. Furthermore, we reveal
the symbol bias hidden in the pairwise comparison of GPT-4o-as-a-Judge and
propose a simple yet effective approach to mitigate it. Experimental results
show that advanced long-context LLMs, such as GPT-4o, perform better in the
many-shot regime than in the zero-shot regime. Meanwhile, the experimental
results further verify the effectiveness of the symbol bias mitigation
approach. | cs.CL | Can Many-Shot In-Context Learning Help Long-Context LLM Judges? See More, Judge Better! |
2024-06-17T15:07:55Z | http://arxiv.org/pdf/2406.11624v1 | http://arxiv.org/abs/2406.11624v1 | Omer Sahin Tas, Royden Wagner | 2024-06-17T15:07:55Z | cs.LG, cs.CL, cs.CV | Motion forecasting transforms sequences of past movements and environment
context into future motion. Recent methods rely on learned representations,
resulting in hidden states that are difficult to interpret. In this work, we
use natural language to quantize motion features in a human-interpretable way,
and measure the degree to which they are embedded in hidden states. Our
experiments reveal that hidden states of motion sequences are arranged with
respect to our discrete sets of motion features. Following these insights, we
fit control vectors to motion features, which allow for controlling motion
forecasts at inference. Consequently, our method enables controlling
transformer-based motion forecasting models with textual inputs, providing a
unique interface to interact with and understand these models. Our
implementation is available at https://github.com/kit-mrt/future-motion | cs.LG | Words in Motion: Representation Engineering for Motion Forecasting |
2024-06-17T15:05:43Z | http://arxiv.org/pdf/2406.11622v1 | http://arxiv.org/abs/2406.11622v1 | Shreya Havaldar, Salvatore Giorgi, Sunny Rai, Thomas Talhelm, Sharath Chandra Guntuku, Lyle Ungar | 2024-06-17T15:05:43Z | cs.CL | Cultural variation exists between nations (e.g., the United States vs.
China), but also within regions (e.g., California vs. Texas, Los Angeles vs.
San Francisco). Measuring this regional cultural variation can illuminate how
and why people think and behave differently. Historically, it has been
difficult to computationally model cultural variation due to a lack of training
data and scalability constraints. In this work, we introduce a new research
problem for the NLP community: How do we measure variation in cultural
constructs across regions using language? We then provide a scalable solution:
building knowledge-guided lexica to model cultural variation, encouraging
future work at the intersection of NLP and cultural understanding. We also
highlight modern LLMs' failure to measure cultural variation or generate
culturally varied language. | cs.CL | Building Knowledge-Guided Lexica to Model Cultural Variation |
2024-06-17T15:02:45Z | http://arxiv.org/pdf/2406.11617v1 | http://arxiv.org/abs/2406.11617v1 | Pala Tej Deep, Rishabh Bhardwaj, Soujanya Poria | 2024-06-17T15:02:45Z | cs.CL | With the proliferation of domain-specific models, model merging has emerged
as a set of techniques that combine the capabilities of multiple models into
one that can multitask without the cost of additional training. In this paper,
we propose a new model merging technique, Drop and rEscaLe via sampLing with
mAgnitude (DELLA-Merging), that employs a novel pruning technique, MAGPRUNE,
which shows significant advantages over DARE and TIES. MAGPRUNE first ranks the
parameters in order of their magnitude and assigns higher dropout probabilities
(p) to parameters with lower ranks corresponding to lower magnitudes. To
approximate the original embeddings, MAGPRUNE employs a rescaling operation on
the parameters that survive the random dropping by 1/(1 - p). On three
different expert models considered for merging (LM, Math, Code) and
corresponding benchmark datasets (AlpacaEval, GSM8K, MBPP), DELLA shows an
average improvement of 2.4 points over baseline methods employing delta
parameter pruning (an improvement of 3.6 points over TIES, 1.2 points over
DARE), and 11.1 points over the no-pruning baseline (TA). We release the source
code at: https://github.com/declare-lab/della. | cs.CL | DELLA-Merging: Reducing Interference in Model Merging through Magnitude-Based Sampling |
2024-06-17T15:00:35Z | http://arxiv.org/pdf/2406.11614v1 | http://arxiv.org/abs/2406.11614v1 | Yihuai Hong, Lei Yu, Shauli Ravfogel, Haiqin Yang, Mor Geva | 2024-06-17T15:00:35Z | cs.CL, cs.AI | The task of "unlearning" certain concepts in large language models (LLMs) has
attracted immense attention recently, due to its importance for mitigating
undesirable model behaviours, such as the generation of harmful, private, or
incorrect information. Current protocols to evaluate unlearning methods largely
rely on behavioral tests, without monitoring the presence of unlearned
knowledge within the model's parameters. This residual knowledge can be
adversarially exploited to recover the erased information post-unlearning. We
argue that unlearning should also be evaluated internally, by considering
changes in the parametric knowledge traces of the unlearned concepts. To this
end, we propose a general methodology for eliciting directions in the parameter
space (termed "concept vectors") that encode concrete concepts, and construct
ConceptVectors, a benchmark dataset containing hundreds of common concepts and
their parametric knowledge traces within two open-source LLMs. Evaluation on
ConceptVectors shows that existing unlearning methods minimally impact concept
vectors, while directly ablating these vectors demonstrably removes the
associated knowledge from the LLMs and significantly reduces their
susceptibility to adversarial manipulation. Our results highlight limitations
in behavioral-based unlearning evaluations and call for future work to include
parametric-based evaluations. To support this, we release our code and
benchmark at https://github.com/yihuaihong/ConceptVectors. | cs.CL | Intrinsic Evaluation of Unlearning Using Parametric Knowledge Traces |
2024-06-17T14:47:06Z | http://arxiv.org/pdf/2406.11598v1 | http://arxiv.org/abs/2406.11598v1 | Arjun Subramonian, Vagrant Gautam, Dietrich Klakow, Zeerak Talat | 2024-06-17T14:47:06Z | cs.CL, cs.CY | Recent improvements in natural language processing (NLP) and machine learning
(ML) and increased mainstream adoption have led to researchers frequently
discussing the "democratization" of artificial intelligence. In this paper, we
seek to clarify how democratization is understood in NLP and ML publications,
through large-scale mixed-methods analyses of papers using the keyword
"democra*" published in NLP and adjacent venues. We find that democratization
is most frequently used to convey (ease of) access to or use of technologies,
without meaningfully engaging with theories of democratization, while research
using other invocations of "democra*" tends to be grounded in theories of
deliberation and debate. Based on our findings, we call for researchers to
enrich their use of the term democratization with appropriate theory, towards
democratic technologies beyond superficial access. | cs.CL | Understanding "Democratization" in NLP and ML Research |
2024-06-17T14:20:53Z | http://arxiv.org/pdf/2406.11581v1 | http://arxiv.org/abs/2406.11581v1 | Shuai Liu, Jonathan May | 2024-06-17T14:20:53Z | cs.CL | Numerous recent techniques for text style transfer characterize their
approaches as variants of reinforcement learning and preference optimization.
In this work, we consider the relationship between these approaches and a class
of optimization approaches developed primarily for (non-neural) statistical
machine translation, formerly known as 'tuning'. Inspired by these techniques
from the past, we improve upon established preference optimization approaches,
incorporating multiple iterations of exploration and optimization, and choosing
contrastive examples by following a 'hope' vs 'fear' sampling strategy.
Cognizant of the difference between machine translation and style transfer,
however, we further tailor our framework with a new pseudo-parallel generation
method and a dynamic weighted reward aggregation method to tackle the lack of
parallel data and the need for a multi-objective reward. We evaluate our model
on two commonly used text style transfer datasets. Through automatic and human
evaluation results we show the effectiveness and the superiority of our model
compared to state-of-the-art baselines. | cs.CL | Style Transfer with Multi-iteration Preference Optimization |
2024-06-17T14:20:47Z | http://arxiv.org/pdf/2406.11580v1 | http://arxiv.org/abs/2406.11580v1 | Tom Kocmi, Vilém Zouhar, Eleftherios Avramidis, Roman Grundkiewicz, Marzena Karpinska, Maja Popović, Mrinmaya Sachan, Mariya Shmatova | 2024-06-17T14:20:47Z | cs.CL | High-quality Machine Translation (MT) evaluation relies heavily on human
judgments. Comprehensive error classification methods, such as Multidimensional
Quality Metrics (MQM), are expensive as they are time-consuming and can only be
done by experts, whose availability may be limited especially for low-resource
languages. On the other hand, just assigning overall scores, like Direct
Assessment (DA), is simpler and faster and can be done by translators of any
level, but are less reliable. In this paper, we introduce Error Span Annotation
(ESA), a human evaluation protocol which combines the continuous rating of DA
with the high-level error severity span marking of MQM. We validate ESA by
comparing it to MQM and DA for 12 MT systems and one human reference
translation (English to German) from WMT23. The results show that ESA offers
faster and cheaper annotations than MQM at the same quality level, without the
requirement of expensive MQM experts. | cs.CL | Error Span Annotation: A Balanced Approach for Human Evaluation of Machine Translation |
2024-06-17T14:11:00Z | http://arxiv.org/pdf/2406.11577v1 | http://arxiv.org/abs/2406.11577v1 | Jacob Collard, Valeria de Paiva, Eswaran Subrahmanian | 2024-06-17T14:11:00Z | cs.CL, math.HO | Mathematics is a highly specialized domain with its own unique set of
challenges. Despite this, there has been relatively little research on natural
language processing for mathematical texts, and there are few mathematical
language resources aimed at NLP. In this paper, we aim to provide annotated
corpora that can be used to study the language of mathematics in different
contexts, ranging from fundamental concepts found in textbooks to advanced
research mathematics. We preprocess the corpora with a neural parsing model and
some manual intervention to provide part-of-speech tags, lemmas, and dependency
trees. In total, we provide 182397 sentences across three corpora. We then aim
to test and evaluate several noteworthy natural language processing models
using these corpora, to show how well they can adapt to the domain of
mathematics and provide useful tools for exploring mathematical language. We
evaluate several neural and symbolic models against benchmarks that we extract
from the corpus metadata to show that terminology extraction and definition
extraction do not easily generalize to mathematics, and that additional work is
needed to achieve good performance on these metrics. Finally, we provide a
learning assistant that grants access to the content of these corpora in a
context-sensitive manner, utilizing text search and entity linking. Though our
corpora and benchmarks provide useful metrics for evaluating mathematical
language processing, further work is necessary to adapt models to mathematics
in order to provide more effective learning assistants and apply NLP methods to
different mathematical domains. | cs.CL | Mathematical Entities: Corpora and Benchmarks |
2024-06-17T14:04:18Z | http://arxiv.org/pdf/2406.11568v1 | http://arxiv.org/abs/2406.11568v1 | Sheng Feng, Heyang Liu, Yu Wang, Yanfeng Wang | 2024-06-17T14:04:18Z | cs.CL, cs.SD, eess.AS, q-bio.NC | In this paper, we introduce a groundbreaking end-to-end (E2E) framework for
decoding invasive brain signals, marking a significant advancement in the field
of speech neuroprosthesis. Our methodology leverages the comprehensive
reasoning abilities of large language models (LLMs) to facilitate direct
decoding. By fully integrating LLMs, we achieve results comparable to the
state-of-the-art cascade models. Our findings underscore the immense potential
of E2E frameworks in speech neuroprosthesis, particularly as the technology
behind brain-computer interfaces (BCIs) and the availability of relevant
datasets continue to evolve. This work not only showcases the efficacy of
combining LLMs with E2E decoding for enhancing speech neuroprosthesis but also
sets a new direction for future research in BCI applications, underscoring the
impact of LLMs in decoding complex neural signals for communication
restoration. Code will be made available at
https://github.com/FsFrancis15/BrainLLM. | cs.CL | Towards an End-to-End Framework for Invasive Brain Signal Decoding with Large Language Models |
2024-06-17T14:03:50Z | http://arxiv.org/pdf/2406.11566v1 | http://arxiv.org/abs/2406.11566v1 | Jiakuan Xie, Pengfei Cao, Yuheng Chen, Yubo Chen, Kang Liu, Jun Zhao | 2024-06-17T14:03:50Z | cs.CL | Knowledge editing aims to adjust the knowledge within large language models
(LLMs) to prevent their responses from becoming obsolete or inaccurate.
However, existing works on knowledge editing are primarily conducted in a
single language, which is inadequate for multilingual language models. In this
paper, we focus on multilingual knowledge editing (MKE), which requires
propagating updates across multiple languages. This necessity poses a
significant challenge for the task. Furthermore, the limited availability of a
comprehensive dataset for MKE exacerbates this challenge, hindering progress in
this area. Hence, we introduce the Multilingual Knowledge Editing Benchmark
(MKEB), a novel dataset comprising 12 languages and providing a complete
evaluation framework. Additionally, we propose a method that enhances
Multilingual knowledge Editing with neuron-Masked Low-Rank Adaptation (MEMLA).
Specifically, we identify two categories of knowledge neurons to improve
editing precision. Moreover, we perform LoRA-based editing with neuron masks to
efficiently modify parameters and facilitate the propagation of updates across
multiple languages. Experiments demonstrate that our method outperforms
existing baselines and significantly enhances the multi-hop reasoning
capability of the edited model, with minimal impact on its downstream task
performance. The dataset and code will be made publicly available. | cs.CL | MEMLA: Enhancing Multilingual Knowledge Editing with Neuron-Masked Low-Rank Adaptation |
2024-06-17T14:03:27Z | http://arxiv.org/pdf/2406.11565v2 | http://arxiv.org/abs/2406.11565v2 | Shaily Bhatt, Fernando Diaz | 2024-06-19T05:18:56Z | cs.CL, cs.CY | Productive interactions between diverse users and language technologies
require outputs from the latter to be culturally relevant and sensitive. Prior
works have evaluated models' knowledge of cultural norms, values, and
artifacts, without considering how this knowledge manifests in downstream
applications. In this work, we focus on extrinsic evaluation of cultural
competence in two text generation tasks, open-ended question answering and
story generation. We quantitatively and qualitatively evaluate model outputs
when an explicit cue of culture, specifically nationality, is perturbed in the
prompts. Although we find that model outputs do vary when varying nationalities
and feature culturally relevant words, we also find weak correlations between
text similarity of outputs for different countries and the cultural values of
these countries. Finally, we discuss important considerations in designing
comprehensive evaluation of cultural competence in user-facing tasks. | cs.CL | Extrinsic Evaluation of Cultural Competence in Large Language Models |
2024-06-17T13:53:15Z | http://arxiv.org/pdf/2406.11555v1 | http://arxiv.org/abs/2406.11555v1 | Lukas Vierling, Jie Fu, Kai Chen | 2024-06-17T13:53:15Z | cs.CL, cs.AI | Recent progress in Large Language Models (LLMs) and language agents has
demonstrated significant promise for various future applications across
multiple disciplines. While traditional approaches to language agents often
rely on fixed, handcrafted designs, our research aims to develop both learnable
and dynamic agents. Our method uses an existing framework that abstracts
language agents as graphs. Within this graph framework, we aim to learn a model
that can generate edges for every given input to the language agent. This
allows us to generate edges that represent the flow of communication within the
graph based on the given input, thereby adjusting the internal communication of
a language agent. We learn to generate these edges using a pretrained LLM that
is fine-tuned with reinforcement learning. This LLM can be fine-tuned on
several datasets simultaneously, and we hypothesize that the model learns to
adapt to these different domains during training, achieving good overall
performance when encountering data from different domains during deployment. We
demonstrate that our approach surpasses the previous static approach by nearly
6% accuracy on a combined dataset of MMLU and CMMLU, and by more than 10% when
trained with a sparsity-inducing loss. It also performs superior in additional
experiments conducted with the MMLU and Mini Crossword Puzzles datasets. The
code is available at https://github.com/lukasVierling/DynamicGPTSwarm. | cs.CL | Input Conditioned Graph Generation for Language Agents |
2024-06-17T13:44:37Z | http://arxiv.org/pdf/2406.11547v1 | http://arxiv.org/abs/2406.11547v1 | Rick Wilming, Artur Dox, Hjalmar Schulz, Marta Oliveira, Benedict Clark, Stefan Haufe | 2024-06-17T13:44:37Z | cs.LG, cs.AI, cs.CL, cs.CY | Large pre-trained language models have become popular for many applications
and form an important backbone of many downstream tasks in natural language
processing (NLP). Applying 'explainable artificial intelligence' (XAI)
techniques to enrich such models' outputs is considered crucial for assuring
their quality and shedding light on their inner workings. However, large
language models are trained on a plethora of data containing a variety of
biases, such as gender biases, affecting model weights and, potentially,
behavior. Currently, it is unclear to what extent such biases also impact model
explanations in possibly unfavorable ways. We create a gender-controlled text
dataset, GECO, in which otherwise identical sentences appear in male and female
forms. This gives rise to ground-truth 'world explanations' for gender
classification tasks, enabling the objective evaluation of the correctness of
XAI methods. We also provide GECOBench, a rigorous quantitative evaluation
framework benchmarking popular XAI methods, applying them to pre-trained
language models fine-tuned to different degrees. This allows us to investigate
how pre-training induces undesirable bias in model explanations and to what
extent fine-tuning can mitigate such explanation bias. We show a clear
dependency between explanation performance and the number of fine-tuned layers,
where XAI methods are observed to particularly benefit from fine-tuning or
complete retraining of embedding layers. Remarkably, this relationship holds
for models achieving similar classification performance on the same task. With
that, we highlight the utility of the proposed gender-controlled dataset and
novel benchmarking approach for research and development of novel XAI methods.
All code including dataset generation, model training, evaluation and
visualization is available at: https://github.com/braindatalab/gecobench | cs.LG | GECOBench: A Gender-Controlled Text Dataset and Benchmark for Quantifying Biases in Explanations |
2024-06-17T13:44:20Z | http://arxiv.org/pdf/2406.11546v1 | http://arxiv.org/abs/2406.11546v1 | Yifan Yang, Zheshu Song, Jianheng Zhuo, Mingyu Cui, Jinpeng Li, Bo Yang, Yexing Du, Ziyang Ma, Xunying Liu, Ziyuan Wang, Ke Li, Shuai Fan, Kai Yu, Wei-Qiang Zhang, Guoguo Chen, Xie Chen | 2024-06-17T13:44:20Z | eess.AS, cs.CL, cs.SD | The evolution of speech technology has been spurred by the rapid increase in
dataset sizes. Traditional speech models generally depend on a large amount of
labeled training data, which is scarce for low-resource languages. This paper
presents GigaSpeech 2, a large-scale, multi-domain, multilingual speech
recognition corpus. It is designed for low-resource languages and does not rely
on paired speech and text data. GigaSpeech 2 comprises about 30,000 hours of
automatically transcribed speech, including Thai, Indonesian, and Vietnamese,
gathered from unlabeled YouTube videos. We also introduce an automated pipeline
for data crawling, transcription, and label refinement. Specifically, this
pipeline uses Whisper for initial transcription and TorchAudio for forced
alignment, combined with multi-dimensional filtering for data quality
assurance. A modified Noisy Student Training is developed to further refine
flawed pseudo labels iteratively, thus enhancing model performance.
Experimental results on our manually transcribed evaluation set and two public
test sets from Common Voice and FLEURS confirm our corpus's high quality and
broad applicability. Notably, ASR models trained on GigaSpeech 2 can reduce the
word error rate for Thai, Indonesian, and Vietnamese on our challenging and
realistic YouTube test set by 25% to 40% compared to the Whisper large-v3
model, with merely 10% model parameters. Furthermore, our ASR models trained on
Gigaspeech 2 yield superior performance compared to commercial services. We
believe that our newly introduced corpus and pipeline will open a new avenue
for low-resource speech recognition and significantly facilitate research in
this area. | eess.AS | GigaSpeech 2: An Evolving, Large-Scale and Multi-domain ASR Corpus for Low-Resource Languages with Automated Crawling, Transcription and Refinement |
2024-06-17T13:21:23Z | http://arxiv.org/pdf/2406.11514v1 | http://arxiv.org/abs/2406.11514v1 | Yi Fang, Moxin Li, Wenjie Wang, Hui Lin, Fuli Feng | 2024-06-17T13:21:23Z | cs.CL | Large Language Models (LLMs) excel in various natural language processing
tasks but struggle with hallucination issues. Existing solutions have
considered utilizing LLMs' inherent reasoning abilities to alleviate
hallucination, such as self-correction and diverse sampling methods. However,
these methods often overtrust LLMs' initial answers due to inherent biases. The
key to alleviating this issue lies in overriding LLMs' inherent biases for
answer inspection. To this end, we propose a CounterFactual Multi-Agent Debate
(CFMAD) framework. CFMAD presets the stances of LLMs to override their inherent
biases by compelling LLMs to generate justifications for a predetermined
answer's correctness. The LLMs with different predetermined stances are engaged
with a skeptical critic for counterfactual debate on the rationality of
generated justifications. Finally, the debate process is evaluated by a
third-party judge to determine the final answer. Extensive experiments on four
datasets of three tasks demonstrate the superiority of CFMAD over existing
methods. | cs.CL | Counterfactual Debating with Preset Stances for Hallucination Elimination of LLMs |
2024-06-17T13:11:17Z | http://arxiv.org/pdf/2406.11930v1 | http://arxiv.org/abs/2406.11930v1 | Abhinav Anand, Shweta Verma, Krishna Narasimhan, Mira Mezini | 2024-06-17T13:11:17Z | cs.SE, cs.AI, cs.CL | Large Language Models trained on code corpora (code-LLMs) have demonstrated
impressive performance in various coding assistance tasks. However, despite
their increased size and training dataset, code-LLMs still have limitations
such as suggesting codes with syntactic errors, variable misuse etc. Some
studies argue that code-LLMs perform well on coding tasks because they use
self-attention and hidden representations to encode relations among input
tokens. However, previous works have not studied what code properties are not
encoded by code-LLMs. In this paper, we conduct a fine-grained analysis of
attention maps and hidden representations of code-LLMs. Our study indicates
that code-LLMs only encode relations among specific subsets of input tokens.
Specifically, by categorizing input tokens into syntactic tokens and
identifiers, we found that models encode relations among syntactic tokens and
among identifiers, but they fail to encode relations between syntactic tokens
and identifiers. We also found that fine-tuned models encode these relations
poorly compared to their pre-trained counterparts. Additionally, larger models
with billions of parameters encode significantly less information about code
than models with only a few hundred million parameters. | cs.SE | A Critical Study of What Code-LLMs (Do Not) Learn |
2024-06-17T13:04:27Z | http://arxiv.org/pdf/2406.11503v1 | http://arxiv.org/abs/2406.11503v1 | Shihao Cai, Keqin Bao, Hangyu Guo, Jizhi Zhang, Jun Song, Bo Zheng | 2024-06-17T13:04:27Z | cs.CV, cs.CL | Large language models have seen widespread adoption in math problem-solving.
However, in geometry problems that usually require visual aids for better
understanding, even the most advanced multi-modal models currently still face
challenges in effectively using image information. High-quality data is crucial
for enhancing the geometric capabilities of multi-modal models, yet existing
open-source datasets and related efforts are either too challenging for direct
model learning or suffer from misalignment between text and images. To overcome
this issue, we introduce a novel pipeline that leverages GPT-4 and GPT-4V to
generate relatively basic geometry problems with aligned text and images,
facilitating model learning. We have produced a dataset of 4.9K geometry
problems and combined it with 19K open-source data to form our GeoGPT4V
dataset. Experimental results demonstrate that the GeoGPT4V dataset
significantly improves the geometry performance of various models on the
MathVista and MathVision benchmarks. The code is available at
https://github.com/Lanyu0303/GeoGPT4V_Project | cs.CV | GeoGPT4V: Towards Geometric Multi-modal Large Language Models with Geometric Image Generation |
2024-06-17T13:01:12Z | http://arxiv.org/pdf/2406.11497v2 | http://arxiv.org/abs/2406.11497v2 | Boyi Deng, Wenjie Wang, Fengbin Zhu, Qifan Wang, Fuli Feng | 2024-06-27T10:18:53Z | cs.CL | Retrieval-Augmented Generation (RAG) can alleviate hallucinations of Large
Language Models (LLMs) by referencing external documents. However, the
misinformation in external documents may mislead LLMs' generation. To address
this issue, we explore the task of "credibility-aware RAG", in which LLMs
automatically adjust the influence of retrieved documents based on their
credibility scores to counteract misinformation. To this end, we introduce a
plug-and-play method named $\textbf{Cr}$edibility-aware $\textbf{A}$ttention
$\textbf{M}$odification (CrAM). CrAM identifies influential attention heads in
LLMs and adjusts their attention weights based on the credibility of the
documents, thereby reducing the impact of low-credibility documents.
Experiments on Natual Questions and TriviaQA using Llama2-13B, Llama3-8B, and
Qwen-7B show that CrAM improves the RAG performance of LLMs against
misinformation pollution by over 20%, even surpassing supervised fine-tuning
methods. | cs.CL | CrAM: Credibility-Aware Attention Modification in LLMs for Combating Misinformation in RAG |
2024-06-17T12:53:21Z | http://arxiv.org/pdf/2406.11486v1 | http://arxiv.org/abs/2406.11486v1 | Vasiliki Kougia, Anastasiia Sedova, Andreas Stephan, Klim Zaporojets, Benjamin Roth | 2024-06-17T12:53:21Z | cs.CL, cs.LG | This paper presents the first study for temporal relation extraction in a
zero-shot setting focusing on biomedical text. We employ two types of prompts
and five LLMs (GPT-3.5, Mixtral, Llama 2, Gemma, and PMC-LLaMA) to obtain
responses about the temporal relations between two events. Our experiments
demonstrate that LLMs struggle in the zero-shot setting performing worse than
fine-tuned specialized models in terms of F1 score, showing that this is a
challenging task for LLMs. We further contribute a novel comprehensive temporal
analysis by calculating consistency scores for each LLM. Our findings reveal
that LLMs face challenges in providing responses consistent to the temporal
properties of uniqueness and transitivity. Moreover, we study the relation
between the temporal consistency of an LLM and its accuracy and whether the
latter can be improved by solving temporal inconsistencies. Our analysis shows
that even when temporal consistency is achieved, the predictions can remain
inaccurate. | cs.CL | Analysing zero-shot temporal relation extraction on clinical notes using temporal consistency |
2024-06-17T12:42:34Z | http://arxiv.org/pdf/2406.11477v1 | http://arxiv.org/abs/2406.11477v1 | Atsuki Yamaguchi, Aline Villavicencio, Nikolaos Aletras | 2024-06-17T12:42:34Z | cs.CL, cs.AI | Large language models (LLMs) have shown remarkable capabilities in many
languages beyond English. Yet, LLMs require more inference steps when
generating non-English text due to their reliance on English-centric
tokenizers, vocabulary, and pre-training data, resulting in higher usage costs
to non-English speakers. Vocabulary expansion with target language tokens is a
widely used cross-lingual vocabulary adaptation approach to remedy this issue.
Despite its effectiveness in inference speedup, the majority of previous work
has focused on high-resource settings assuming access to a substantial amount
of target language data to effectively initialize the embeddings of the new
tokens and adapt the LLM to the target language. However, vocabulary expansion
for LLMs in low-resource settings (i.e. languages and compute) has yet to be
explored. In this paper, we investigate sample-efficient adaptation strategies
from different angles, including target vocabulary size and initialization
methods, and the amount of target data available for adaptation. Extensive
experiments across typologically diverse languages, tasks and models show that
simpler heuristic-based embedding initialization is more efficient and robust
to changes in target vocabulary size and adaptation data in low-resource
settings, outperforming a popular random initialization and a more
sophisticated state-of-the-art approach that relies on external data and model. | cs.CL | Vocabulary Expansion for Low-resource Cross-lingual Transfer |
2024-06-17T12:38:48Z | http://arxiv.org/pdf/2406.11474v1 | http://arxiv.org/abs/2406.11474v1 | Heyan Huang, Yinghao Li, Huashan Sun, Yu Bai, Yang Gao | 2024-06-17T12:38:48Z | cs.CL, cs.AI | Recent studies have demonstrated that In-Context Learning (ICL), through the
use of specific demonstrations, can align Large Language Models (LLMs) with
human preferences known as In-Context Alignment (ICA), indicating that models
can comprehend human instructions without requiring parameter adjustments.
However, the exploration of the mechanism and applicability of ICA remains
limited. In this paper, we begin by dividing the context text used in ICA into
three categories: format, system prompt, and example. Through ablation
experiments, we investigate the effectiveness of each part in enabling ICA to
function effectively. We then examine how variants in these parts impact the
model's alignment performance. Our findings indicate that the example part is
crucial for enhancing the model's alignment capabilities, with changes in
examples significantly affecting alignment performance. We also conduct a
comprehensive evaluation of ICA's zero-shot capabilities in various alignment
tasks. The results indicate that compared to parameter fine-tuning methods, ICA
demonstrates superior performance in knowledge-based tasks and tool-use tasks.
However, it still exhibits certain limitations in areas such as multi-turn
dialogues and instruction following. | cs.CL | How Far Can In-Context Alignment Go? Exploring the State of In-Context Alignment |
2024-06-17T12:38:38Z | http://arxiv.org/pdf/2406.11473v1 | http://arxiv.org/abs/2406.11473v1 | Justin Deschenaux, Caglar Gulcehre | 2024-06-17T12:38:38Z | cs.CL, cs.AI | The modern autoregressive Large Language Models (LLMs) have achieved
outstanding performance on NLP benchmarks, and they are deployed in the real
world. However, they still suffer from limitations of the autoregressive
training paradigm. For example, autoregressive token generation is notably slow
and can be prone to \textit{exposure bias}. The diffusion-based language models
were proposed as an alternative to autoregressive generation to address some of
these limitations. We evaluate the recently proposed Score Entropy Discrete
Diffusion (SEDD) approach and show it is a promising alternative to
autoregressive generation but it has some short-comings too. We empirically
demonstrate the advantages and challenges of SEDD, and observe that SEDD
generally matches autoregressive models in perplexity and on benchmarks such as
HellaSwag, Arc or WinoGrande. Additionally, we show that in terms of inference
latency, SEDD can be up to 4.5$\times$ more efficient than GPT-2. While SEDD
allows conditioning on tokens at abitrary positions, SEDD appears slightly
weaker than GPT-2 for conditional generation given short prompts. Finally, we
reproduced the main results from the original SEDD paper. | cs.CL | Promises, Outlooks and Challenges of Diffusion Language Modeling |
2024-06-17T12:25:25Z | http://arxiv.org/pdf/2406.11464v1 | http://arxiv.org/abs/2406.11464v1 | Jesús Calleja, Thierry Etchegoyhen, David Ponce | 2024-06-17T12:25:25Z | cs.CL | Easy Read text is one of the main forms of access to information for people
with reading difficulties. One of the key characteristics of this type of text
is the requirement to split sentences into smaller grammatical segments, to
facilitate reading. Automated segmentation methods could foster the creation of
Easy Read content, but their viability has yet to be addressed. In this work,
we study novel methods for the task, leveraging masked and generative language
models, along with constituent parsing. We conduct comprehensive automatic and
human evaluations in three languages, analysing the strengths and weaknesses of
the proposed alternatives, under scarce resource limitations. Our results
highlight the viability of automated ER segmentation and remaining deficiencies
compared to expert-driven human segmentation. | cs.CL | Automating Easy Read Text Segmentation |
2024-06-17T12:23:32Z | http://arxiv.org/pdf/2406.11460v1 | http://arxiv.org/abs/2406.11460v1 | Jinyuan Fang, Zaiqiao Meng, Craig Macdonald | 2024-06-17T12:23:32Z | cs.CL, cs.AI | Retrieval-augmented generation (RAG) offers an effective approach for
addressing question answering (QA) tasks. However, the imperfections of the
retrievers in RAG models often result in the retrieval of irrelevant
information, which could introduce noises and degrade the performance,
especially when handling multi-hop questions that require multiple steps of
reasoning. To enhance the multi-hop reasoning ability of RAG models, we propose
TRACE. TRACE constructs knowledge-grounded reasoning chains, which are a series
of logically connected knowledge triples, to identify and integrate supporting
evidence from the retrieved documents for answering questions. Specifically,
TRACE employs a KG Generator to create a knowledge graph (KG) from the
retrieved documents, and then uses an Autoregressive Reasoning Chain
Constructor to build reasoning chains. Experimental results on three multi-hop
QA datasets show that TRACE achieves an average performance improvement of up
to 14.03% compared to using all the retrieved documents. Moreover, the results
indicate that using reasoning chains as context, rather than the entire
documents, is often sufficient to correctly answer questions. | cs.CL | TRACE the Evidence: Constructing Knowledge-Grounded Reasoning Chains for Retrieval-Augmented Generation |
2024-06-17T12:11:01Z | http://arxiv.org/pdf/2406.11455v1 | http://arxiv.org/abs/2406.11455v1 | Zepeng Ding, Ruiyang Ke, Wenhao Huang, Guochao Jiang, Yanda Li, Deqing Yang, Yanghua Xiao, Jiaqing Liang | 2024-06-17T12:11:01Z | cs.CL, cs.AI | Existing research on large language models (LLMs) shows that they can solve
information extraction tasks through multi-step planning. However, their
extraction behavior on complex sentences and tasks is unstable, emerging issues
such as false positives and missing elements. We observe that decomposing
complex extraction tasks and extracting them step by step can effectively
improve LLMs' performance, and the extraction orders of entities significantly
affect the final results of LLMs. This paper proposes a two-stage multi-step
method for LLM-based information extraction and adopts the RL framework to
execute the multi-step planning. We regard sequential extraction as a Markov
decision process, build an LLM-based extraction environment, design a decision
module to adaptively provide the optimal order for sequential entity extraction
on different sentences, and utilize the DDQN algorithm to train the decision
model. We also design the rewards and evaluation metrics suitable for the
extraction results of LLMs. We conduct extensive experiments on multiple public
datasets to demonstrate the effectiveness of our method in improving the
information extraction capabilities of LLMs. | cs.CL | Adaptive Reinforcement Learning Planning: Harnessing Large Language Models for Complex Information Extraction |
2024-06-17T11:36:39Z | http://arxiv.org/pdf/2406.11431v1 | http://arxiv.org/abs/2406.11431v1 | Wenkai Yang, Shiqi Shen, Guangyao Shen, Zhi Gong, Yankai Lin | 2024-06-17T11:36:39Z | cs.CL, cs.AI | Superalignment, where humans are weak supervisors of superhuman models, has
become an important and widely discussed issue in the current era of rapid
development of Large Language Models (LLMs). The recent work preliminarily
studies this problem by using weak models to supervise strong models. It
discovers that weakly supervised strong students can consistently outperform
weak teachers towards the alignment target, leading to a weak-to-strong
generalization phenomenon. However, we are concerned that behind such a
promising phenomenon, whether there exists an issue of weak-to-strong
deception, where strong models may deceive weak models by exhibiting
well-aligned in areas known to weak models but producing misaligned behaviors
in cases weak models do not know. We then take an initial step towards
exploring this security issue in a specific but realistic multi-objective
alignment case, where there may be some alignment targets conflicting with each
other (e.g., helpfulness v.s. harmlessness). Such a conflict is likely to cause
strong models to deceive weak models in one alignment dimension to gain high
reward in other alignment dimension. Our experiments on both the reward
modeling task and the preference optimization scenario indicate: (1) the
weak-to-strong deception exists; (2) the deception phenomenon may intensify as
the capability gap between weak and strong models increases. We also discuss
potential solutions and find bootstrapping with an intermediate model can
mitigate the deception to some extent. Our work highlights the urgent need to
pay more attention to the true reliability of superalignment. | cs.CL | Super(ficial)-alignment: Strong Models May Deceive Weak Models in Weak-to-Strong Generalization |
2024-06-17T11:35:16Z | http://arxiv.org/pdf/2406.11430v1 | http://arxiv.org/abs/2406.11430v1 | Alessio Devoto, Yu Zhao, Simone Scardapane, Pasquale Minervini | 2024-06-17T11:35:16Z | cs.CL, cs.AI | The deployment of large language models (LLMs) is often hindered by the
extensive memory requirements of the Key-Value (KV) cache, especially as
context lengths increase. Existing approaches to reduce the KV cache size
involve either fine-tuning the model to learn a compression strategy or
leveraging attention scores to reduce the sequence length. We analyse the
attention distributions in decoder-only Transformers-based models and observe
that attention allocation patterns stay consistent across most layers.
Surprisingly, we find a clear correlation between the $L_2$ and the attention
scores over cached KV pairs, where a low $L_2$ of a key embedding usually leads
to a high attention score during decoding. This finding indicates that the
influence of a KV pair is potentially determined by the key embedding itself
before being queried. Based on this observation, we compress the KV cache based
on the $L_2$ of key embeddings. Our experimental results show that this simple
strategy can reduce the KV cache size by 50% on language modelling and
needle-in-a-haystack tasks and 90% on passkey retrieval tasks without losing
accuracy. | cs.CL | A Simple and Effective $L_2$ Norm-Based Strategy for KV Cache Compression |
2024-06-17T11:31:48Z | http://arxiv.org/pdf/2406.11429v1 | http://arxiv.org/abs/2406.11429v1 | Shilong Li, Ge Bai, Zhang Zhang, Ying Liu, Chenji Lu, Daichi Guo, Ruifang Liu, Yong Sun | 2024-06-17T11:31:48Z | cs.CL, cs.AI | Predicting unseen relations that cannot be observed during the training phase
is a challenging task in relation extraction. Previous works have made progress
by matching the semantics between input instances and label descriptions.
However, fine-grained matching often requires laborious manual annotation, and
rich interactions between instances and label descriptions come with
significant computational overhead. In this work, we propose an efficient
multi-grained matching approach that uses virtual entity matching to reduce
manual annotation cost, and fuses coarse-grained recall and fine-grained
classification for rich interactions with guaranteed inference speed.
Experimental results show that our approach outperforms the previous State Of
The Art (SOTA) methods, and achieves a balance between inference efficiency and
prediction accuracy in zero-shot relation extraction tasks. Our code is
available at https://github.com/longls777/EMMA. | cs.CL | Fusion Makes Perfection: An Efficient Multi-Grained Matching Approach for Zero-Shot Relation Extraction |
2024-06-17T11:25:57Z | http://arxiv.org/pdf/2406.11427v1 | http://arxiv.org/abs/2406.11427v1 | Keon Lee, Dong Won Kim, Jaehyeon Kim, Jaewoong Cho | 2024-06-17T11:25:57Z | eess.AS, cs.AI, cs.CL, cs.LG, cs.SD | Large-scale diffusion models have shown outstanding generative abilities
across multiple modalities including images, videos, and audio. However,
text-to-speech (TTS) systems typically involve domain-specific modeling factors
(e.g., phonemes and phoneme-level durations) to ensure precise temporal
alignments between text and speech, which hinders the efficiency and
scalability of diffusion models for TTS. In this work, we present an efficient
and scalable Diffusion Transformer (DiT) that utilizes off-the-shelf
pre-trained text and speech encoders. Our approach addresses the challenge of
text-speech alignment via cross-attention mechanisms with the prediction of the
total length of speech representations. To achieve this, we enhance the DiT
architecture to suit TTS and improve the alignment by incorporating semantic
guidance into the latent space of speech. We scale the training dataset and the
model size to 82K hours and 790M parameters, respectively. Our extensive
experiments demonstrate that the large-scale diffusion model for TTS without
domain-specific modeling not only simplifies the training pipeline but also
yields superior or comparable zero-shot performance to state-of-the-art TTS
models in terms of naturalness, intelligibility, and speaker similarity. Our
speech samples are available at https://ditto-tts.github.io. | eess.AS | DiTTo-TTS: Efficient and Scalable Zero-Shot Text-to-Speech with Diffusion Transformer |
2024-06-17T11:22:25Z | http://arxiv.org/pdf/2406.11424v1 | http://arxiv.org/abs/2406.11424v1 | Gautam B, Anupam Purwar | 2024-06-17T11:22:25Z | cs.IR, cs.CL | This paper presents an analysis of open-source large language models (LLMs)
and their application in Retrieval-Augmented Generation (RAG) tasks, specific
for enterprise-specific data sets scraped from their websites. With the
increasing reliance on LLMs in natural language processing, it is crucial to
evaluate their performance, accessibility, and integration within specific
organizational contexts. This study examines various open-source LLMs, explores
their integration into RAG frameworks using enterprise-specific data, and
assesses the performance of different open-source embeddings in enhancing the
retrieval and generation process. Our findings indicate that open-source LLMs,
combined with effective embedding techniques, can significantly improve the
accuracy and efficiency of RAG systems, offering a viable alternative to
proprietary solutions for enterprises. | cs.IR | Evaluating the Efficacy of Open-Source LLMs in Enterprise-Specific RAG Systems: A Comparative Study of Performance and Scalability |
2024-06-17T11:22:04Z | http://arxiv.org/pdf/2406.11423v1 | http://arxiv.org/abs/2406.11423v1 | Evan M. Williams, Peter Carragher, Kathleen M. Carley | 2024-06-17T11:22:04Z | cs.SI, cs.AI, cs.CL, cs.CY, cs.LG | In an attempt to mimic the complex paths through which unreliable content
spreads between search engines and social media, we explore the impact of
incorporating both webgraph and large-scale social media contexts into website
credibility classification and discovery systems. We further explore the usage
of what we define as \textit{dredge words} on social media -- terms or phrases
for which unreliable domains rank highly. Through comprehensive graph neural
network ablations, we demonstrate that curriculum-based heterogeneous graph
models that leverage context from both webgraphs and social media data
outperform homogeneous and single-mode approaches. We further demonstrate that
the incorporation of dredge words into our model strongly associates unreliable
websites with social media and online commerce platforms. Finally, we show our
heterogeneous model greatly outperforms competing systems in the top-k
identification of unlabeled unreliable websites. We demonstrate the strong
unreliability signals present in the diverse paths that users follow to uncover
unreliable content, and we release a novel dataset of dredge words. | cs.SI | Dredge Word, Social Media, and Webgraph Networks for Unreliable Website Classification and Identification |
2024-06-17T11:08:08Z | http://arxiv.org/pdf/2406.11418v1 | http://arxiv.org/abs/2406.11418v1 | Zhewen Shen, Aditya Joshi, Ruey-Cheng Chen | 2024-06-17T11:08:08Z | cs.CL | Children from bilingual backgrounds benefit from interactions with parents
and teachers to re-acquire their heritage language. In this paper, we
investigate how this insight from behavioral study can be incorporated into the
learning of small-scale language models. We introduce BAMBINO-LM, a continual
pretraining strategy for BabyLM that uses a novel combination of alternation
and PPO-based perplexity reward induced from a parent Italian model. Upon
evaluation on zero-shot classification tasks for English and Italian,
BAMBINO-LM improves the Italian language capability of a BabyLM baseline. Our
ablation analysis demonstrates that employing both the alternation strategy and
PPO-based modeling is key to this effectiveness gain. We also show that, as a
side effect, the proposed method leads to similar degradation in L1
effectiveness as human children would have had in an equivalent learning
scenario. | cs.CL | BAMBINO-LM: (Bilingual-)Human-Inspired Continual Pretraining of BabyLM |
2024-06-17T11:05:15Z | http://arxiv.org/pdf/2406.15486v2 | http://arxiv.org/abs/2406.15486v2 | Qianchao Zhu, Jiangfei Duan, Chang Chen, Siran Liu, Xiuhong Li, Guanyu Feng, Xin Lv, Huanqi Cao, Xiao Chuanfu, Xingcheng Zhang, Dahua Lin, Chao Yang | 2024-06-28T08:55:17Z | cs.CL, cs.AI, cs.LG | Large language models (LLMs) now support extremely long context windows, but
the quadratic complexity of vanilla attention results in significantly long
Time-to-First-Token (TTFT) latency. Existing approaches to address this
complexity require additional pretraining or finetuning, and often sacrifice
model accuracy. In this paper, we first provide both theoretical and empirical
foundations for near-lossless sparse attention. We find dynamically capturing
head-specific sparse patterns at runtime with low overhead is crucial. To
address this, we propose SampleAttention, an adaptive structured and
near-lossless sparse attention. Leveraging observed significant sparse
patterns, SampleAttention attends to a fixed percentage of adjacent tokens to
capture local window patterns, and employs a two-stage query-guided key-value
filtering approach, which adaptively select a minimum set of key-values with
low overhead, to capture column stripe patterns. Comprehensive evaluations show
that SampleAttention can seamlessly replace vanilla attention in off-the-shelf
LLMs with nearly no accuracy loss, and reduces TTFT by up to $2.42\times$
compared with FlashAttention. | cs.CL | SampleAttention: Near-Lossless Acceleration of Long Context LLM Inference with Adaptive Structured Sparse Attention |
2024-06-17T11:00:04Z | http://arxiv.org/pdf/2406.15485v2 | http://arxiv.org/abs/2406.15485v2 | Xingjian Hu, Baole Wei, Liangcai Gao | 2024-06-25T07:37:22Z | cs.CL, cs.CV | Text line detection is a key task in historical document analysis facing many
challenges of arbitrary-shaped text lines, dense texts, and text lines with
high aspect ratios, etc. In this paper, we propose a general framework for
historical document text detection (SegHist), enabling existing
segmentation-based text detection methods to effectively address the
challenges, especially text lines with high aspect ratios. Integrating the
SegHist framework with the commonly used method DB++, we develop DB-SegHist.
This approach achieves SOTA on the CHDAC, MTHv2, and competitive results on
HDRC datasets, with a significant improvement of 1.19% on the most challenging
CHDAC dataset which features more text lines with high aspect ratios. Moreover,
our method attains SOTA on rotated MTHv2 and rotated HDRC, demonstrating its
rotational robustness. The code is available at
https://github.com/LumionHXJ/SegHist. | cs.CL | SegHist: A General Segmentation-based Framework for Chinese Historical Document Text Line Detection |
2024-06-17T10:56:03Z | http://arxiv.org/pdf/2406.11410v2 | http://arxiv.org/abs/2406.11410v2 | Lingyun Zhang, Bin jin, Gaojian Ge, Lunhui Liu, Xuewen Shen, Mingyong Wu, Houqian Zhang, Yongneng Jiang, Shiqi Chen, Shi Pu | 2024-06-18T11:59:03Z | cs.CL, cs.AI | Human priors play a crucial role in efficiently utilizing data in deep
learning. However, with the development of large language models (LLMs), there
is an increasing emphasis on scaling both model size and data volume, which
often diminishes the importance of human priors in data construction.
Influenced by these trends, existing Small Language Models (SLMs) mainly rely
on web-scraped large-scale training data, neglecting the proper incorporation
of human priors. This oversight limits the training efficiency of language
models in resource-constrained settings. In this paper, we propose a principle
to leverage human priors for data construction. This principle emphasizes
achieving high-performance SLMs by training on a concise dataset that
accommodates both semantic diversity and data quality consistency, while
avoiding benchmark data leakage. Following this principle, we train an SLM
named HARE-1.1B. Extensive experiments on large-scale benchmark datasets
demonstrate that HARE-1.1B performs favorably against state-of-the-art SLMs,
validating the effectiveness of the proposed principle. Additionally, this
provides new insights into efficient language model training in
resource-constrained environments from the view of human priors. | cs.CL | HARE: HumAn pRiors, a key to small language model Efficiency |
2024-06-17T10:54:35Z | http://arxiv.org/pdf/2406.11409v2 | http://arxiv.org/abs/2406.11409v2 | CodeGemma Team, Heri Zhao, Jeffrey Hui, Joshua Howland, Nam Nguyen, Siqi Zuo, Andrea Hu, Christopher A. Choquette-Choo, Jingyue Shen, Joe Kelley, Kshitij Bansal, Luke Vilnis, Mateo Wirth, Paul Michel, Peter Choy, Pratik Joshi, Ravin Kumar, Sarmad Hashmi, Shubham Agrawal, Zhitao Gong, Jane Fine, Tris Warkentin, Ale Jakse Hartman, Bin Ni, Kathy Korevec, Kelly Schaefer, Scott Huffman | 2024-06-19T02:37:50Z | cs.CL, cs.AI | This paper introduces CodeGemma, a collection of specialized open code models
built on top of Gemma, capable of a variety of code and natural language
generation tasks. We release three model variants. CodeGemma 7B pretrained (PT)
and instruction-tuned (IT) variants have remarkably resilient natural language
understanding, excel in mathematical reasoning, and match code capabilities of
other open models. CodeGemma 2B is a state-of-the-art code completion model
designed for fast code infilling and open-ended generation in latency-sensitive
settings. | cs.CL | CodeGemma: Open Code Models Based on Gemma |
2024-06-17T10:45:47Z | http://arxiv.org/pdf/2406.11403v1 | http://arxiv.org/abs/2406.11403v1 | Franz Louis Cesista | 2024-06-17T10:45:47Z | cs.CV, cs.CL | Multimodal Foundation Models (MMFMs) have shown remarkable performance on
various computer vision and natural language processing tasks. However, their
performance on particular tasks such as document understanding is still
limited. They also require more compute, time, and engineering resources to
finetune and deploy compared to traditional, unimodal models. In this report,
we present Multimodal Structured Generation, a general framework which
constrains the output logits of frozen MMFMs to force them to reason before
responding with structured outputs that downstream APIs can parse and use. We
provide a detailed account of our approach, including the technical details,
theoretical discussions, and final evaluation results in the 2nd Multimodal
Foundation Models Challenge hosted by the Computer Vision and Pattern
Recognition (CVPR) conference. Our approach achieved the second highest score
in the hidden test set for Phase 2 and third highest overall. This shows the
method's ability to generalize to unseen tasks. And that simple engineering can
beat expensive & complicated modelling steps as we first discussed in our
paper, Retrieval Augmented Structured Generation: Business Document Information
Extraction as Tool Use. All of our scripts, deployment steps, and evaluation
results can be accessed in https://github.com/leloykun/MMFM-Challenge | cs.CV | Multimodal Structured Generation: CVPR's 2nd MMFM Challenge Technical Report |
2024-06-17T10:45:36Z | http://arxiv.org/pdf/2406.11402v1 | http://arxiv.org/abs/2406.11402v1 | Neelabh Sinha, Vinija Jain, Aman Chadha | 2024-06-17T10:45:36Z | cs.CL, cs.AI, cs.LG | The rapid rise of Language Models (LMs) has expanded their use in several
applications. Yet, due to constraints of model size, associated cost, or
proprietary restrictions, utilizing state-of-the-art (SOTA) LLMs is not always
feasible. With open, smaller LMs emerging, more applications can leverage their
capabilities, but selecting the right LM can be challenging. This work conducts
an in-depth experimental analysis of the semantic correctness of outputs of 10
smaller, open LMs across three aspects: task types, application domains and
reasoning types, using diverse prompt styles. We demonstrate that most
effective models and prompt styles vary depending on the specific requirements.
Our analysis provides a comparative assessment of LMs and prompt styles using a
proposed three-tier schema of aspects for their strategic selection based on
use-case and other constraints. We also show that if utilized appropriately,
these LMs can compete with, and sometimes outperform, SOTA LLMs like
DeepSeek-v2, GPT-3.5-Turbo, and GPT-4o. | cs.CL | Evaluating Open Language Models Across Task Types, Application Domains, and Reasoning Types: An In-Depth Experimental Analysis |
2024-06-17T10:38:03Z | http://arxiv.org/pdf/2406.11400v1 | http://arxiv.org/abs/2406.11400v1 | Golnaz Shapurian | 2024-06-17T10:38:03Z | cs.CL, astro-ph.IM | This paper presents an experiment conducted during a hackathon, focusing on
using large language models (LLMs) and knowledge graph clustering to extract
entities and relationships from astronomical text. The study demonstrates an
approach to disambiguate entities that can appear in various contexts within
the astronomical domain. By collecting excerpts around specific entities and
leveraging the GPT-4 language model, relevant entities and relationships are
extracted. The extracted information is then used to construct a knowledge
graph, which is clustered using the Leiden algorithm. The resulting Leiden
communities are utilized to identify the percentage of association of unknown
excerpts to each community, thereby enabling disambiguation. The experiment
showcases the potential of combining LLMs and knowledge graph clustering
techniques for information extraction in astronomical research. The results
highlight the effectiveness of the approach in identifying and disambiguating
entities, as well as grouping them into meaningful clusters based on their
relationships. | cs.CL | Large Language Models and Knowledge Graphs for Astronomical Entity Disambiguation |
2024-06-17T10:12:45Z | http://arxiv.org/pdf/2406.11385v2 | http://arxiv.org/abs/2406.11385v2 | Yuyan Zhou, Liang Song, Bingning Wang, Weipeng Chen | 2024-06-27T16:01:28Z | cs.CL | The advent of large language models (LLMs) like GPT-4 has catalyzed the
exploration of multi-task learning (MTL), in which a single model demonstrates
proficiency across diverse tasks. Task arithmetic has emerged as a
cost-effective approach for MTL. It enables performance enhancement across
multiple tasks by adding their corresponding task vectors to a pre-trained
model. However, the current lack of a method that can simultaneously achieve
optimal performance, computational efficiency, and data privacy limits their
application to LLMs. In this paper, we propose \textbf{M}odel
\textbf{E}xclusive \textbf{T}ask \textbf{A}rithmetic for merging
\textbf{GPT}-scale models, which formalizes the objective of model merging into
a multi-task learning framework, aiming to minimize the average loss difference
between the merged model and each individual task model. Since data privacy
limits the use of multi-task training data, we leverage LLMs' local linearity
and task vectors' orthogonality to separate the data term and scaling
coefficients term and derive a model-exclusive task arithmetic method. Our
proposed MetaGPT is data-agnostic and bypasses the heavy search process, making
it cost-effective and easy to implement for LLMs.Extensive experiments
demonstrate that MetaGPT leads to improvements in task arithmetic and achieves
state-of-the-art performance on multiple tasks. | cs.CL | MetaGPT: Merging Large Language Models Using Model Exclusive Task Arithmetic |
2024-06-17T09:56:46Z | http://arxiv.org/pdf/2406.11380v1 | http://arxiv.org/abs/2406.11380v1 | Gaspard Michel, Elena V. Epure, Romain Hennequin, Christophe Cerisara | 2024-06-17T09:56:46Z | cs.CL | Large Language Models (LLMs) zero-shot and few-shot performance are subject
to memorization and data contamination, complicating the assessment of their
validity. In literary tasks, the performance of LLMs is often correlated to the
degree of book memorization. In this work, we carry out a realistic evaluation
of LLMs for quotation attribution in novels, taking the instruction fined-tuned
version of Llama3 as an example. We design a task-specific memorization measure
and use it to show that Llama3's ability to perform quotation attribution is
positively correlated to the novel degree of memorization. However, Llama3
still performs impressively well on books it has not memorized nor seen. Data
and code will be made publicly available. | cs.CL | A Realistic Evaluation of LLMs for Quotation Attribution in Literary Texts: A Case Study of LLaMa3 |
2024-06-17T09:51:38Z | http://arxiv.org/pdf/2406.11375v1 | http://arxiv.org/abs/2406.11375v1 | Siyu Yuan, Cheng Jiayang, Lin Qiu, Deqing Yang | 2024-06-17T09:51:38Z | cs.CL, cs.AI | Analogical reasoning plays a critical role in human cognition, enabling us to
understand new concepts by associating them with familiar ones. Previous
research in the AI community has mainly focused on identifying and generating
analogies and then examining their quality under human evaluation, which
overlooks the practical application of these analogies in real-world settings.
Inspired by the human education process, in this paper, we propose to
investigate how analogies created by teacher language models (LMs) can assist
student LMs in understanding scientific concepts, thereby aligning more closely
with practical scenarios. Our results suggest that free-form analogies can
indeed aid LMs in understanding concepts. Additionally, analogies generated by
student LMs can improve their own performance on scientific question answering,
demonstrating their capability to use analogies for self-learning new
knowledge. Resources are available at https://github.com/siyuyuan/SCUA. | cs.CL | Boosting Scientific Concepts Understanding: Can Analogy from Teacher Models Empower Student Models? |
2024-06-17T09:48:53Z | http://arxiv.org/pdf/2406.11370v1 | http://arxiv.org/abs/2406.11370v1 | Han Zhou, Xingchen Wan, Yinhong Liu, Nigel Collier, Ivan Vulić, Anna Korhonen | 2024-06-17T09:48:53Z | cs.CL, cs.AI, cs.CY, cs.LG | Large language models (LLMs) have shown promising abilities as cost-effective
and reference-free evaluators for assessing language generation quality. In
particular, pairwise LLM evaluators, which compare two generated texts and
determine the preferred one, have been employed in a wide range of
applications. However, LLMs exhibit preference biases and worrying sensitivity
to prompt designs. In this work, we first reveal that the predictive preference
of LLMs can be highly brittle and skewed, even with semantically equivalent
instructions. We find that fairer predictive preferences from LLMs consistently
lead to judgments that are better aligned with humans. Motivated by this
phenomenon, we propose an automatic Zero-shot Evaluation-oriented Prompt
Optimization framework, ZEPO, which aims to produce fairer preference decisions
and improve the alignment of LLM evaluators with human judgments. To this end,
we propose a zero-shot learning objective based on the preference decision
fairness. ZEPO demonstrates substantial performance improvements over
state-of-the-art LLM evaluators, without requiring labeled data, on
representative meta-evaluation benchmarks. Our findings underscore the critical
correlation between preference fairness and human alignment, positioning ZEPO
as an efficient prompt optimizer for bridging the gap between LLM evaluators
and human judgments. | cs.CL | Fairer Preferences Elicit Improved Human-Aligned Large Language Model Judgments |
2024-06-17T09:46:35Z | http://arxiv.org/pdf/2406.11368v1 | http://arxiv.org/abs/2406.11368v1 | Gaspard Michel, Elena V. Epure, Romain Hennequin, Christophe Cerisara | 2024-06-17T09:46:35Z | cs.CL | Humans naturally attribute utterances of direct speech to their speaker in
literary works. When attributing quotes, we process contextual information but
also access mental representations of characters that we build and revise
throughout the narrative. Recent methods to automatically attribute such
utterances have explored simulating human logic with deterministic rules or
learning new implicit rules with neural networks when processing contextual
information. However, these systems inherently lack \textit{character}
representations, which often leads to errors on more challenging examples of
attribution: anaphoric and implicit quotes. In this work, we propose to augment
a popular quotation attribution system, BookNLP, with character embeddings that
encode global information of characters. To build these embeddings, we create
DramaCV, a corpus of English drama plays from the 15th to 20th century focused
on Character Verification (CV), a task similar to Authorship Verification (AV),
that aims at analyzing fictional characters. We train a model similar to the
recently proposed AV model, Universal Authorship Representation (UAR), on this
dataset, showing that it outperforms concurrent methods of characters
embeddings on the CV task and generalizes better to literary novels. Then,
through an extensive evaluation on 22 novels, we show that combining BookNLP's
contextual information with our proposed global character embeddings improves
the identification of speakers for anaphoric and implicit quotes, reaching
state-of-the-art performance. Code and data will be made publicly available. | cs.CL | Improving Quotation Attribution with Fictional Character Embeddings |
2024-06-17T09:39:34Z | http://arxiv.org/pdf/2406.16937v1 | http://arxiv.org/abs/2406.16937v1 | Sumit Kumar Dam, Choong Seon Hong, Yu Qiao, Chaoning Zhang | 2024-06-17T09:39:34Z | cs.CL, cs.AI | The past few decades have witnessed an upsurge in data, forming the
foundation for data-hungry, learning-based AI technology. Conversational
agents, often referred to as AI chatbots, rely heavily on such data to train
large language models (LLMs) and generate new content (knowledge) in response
to user prompts. With the advent of OpenAI's ChatGPT, LLM-based chatbots have
set new standards in the AI community. This paper presents a complete survey of
the evolution and deployment of LLM-based chatbots in various sectors. We first
summarize the development of foundational chatbots, followed by the evolution
of LLMs, and then provide an overview of LLM-based chatbots currently in use
and those in the development phase. Recognizing AI chatbots as tools for
generating new knowledge, we explore their diverse applications across various
industries. We then discuss the open challenges, considering how the data used
to train the LLMs and the misuse of the generated knowledge can cause several
issues. Finally, we explore the future outlook to augment their efficiency and
reliability in numerous applications. By addressing key milestones and the
present-day context of LLM-based chatbots, our survey invites readers to delve
deeper into this realm, reflecting on how their next generation will reshape
conversational AI. | cs.CL | A Complete Survey on LLM-based AI Chatbots |
2024-06-17T09:25:10Z | http://arxiv.org/pdf/2406.11357v2 | http://arxiv.org/abs/2406.11357v2 | Zhonghao Li, Xuming Hu, Aiwei Liu, Kening Zheng, Sirui Huang, Hui Xiong | 2024-06-18T02:44:27Z | cs.CL, cs.AI | Large Language Models (LLMs) are limited by their parametric knowledge,
leading to hallucinations in knowledge-extensive tasks. To address this,
Retrieval-Augmented Generation (RAG) incorporates external document chunks to
expand LLM knowledge. Furthermore, compressing information from document chunks
through extraction or summarization can improve LLM performance. Nonetheless,
LLMs still struggle to notice and utilize scattered key information, a problem
known as the "lost-in-the-middle" syndrome. Therefore, we typically need to
restructure the content for LLM to recognize the key information. We propose
$\textit{Refiner}$, an end-to-end extract-and-restructure paradigm that
operates in the post-retrieval process of RAG. $\textit{Refiner}$ leverages a
single decoder-only LLM to adaptively extract query-relevant contents verbatim
along with the necessary context, and section them based on their
interconnectedness, thereby highlights information distinction, and aligns
downstream LLMs with the original context effectively. Experiments show that a
trained $\textit{Refiner}$ (with 7B parameters) exhibits significant gain to
downstream LLM in improving answer accuracy, and outperforms other
state-of-the-art advanced RAG and concurrent compressing approaches in various
single-hop and multi-hop QA tasks. Notably, $\textit{Refiner}$ achieves a 80.5%
tokens reduction and a 1.6-7.0% improvement margin in multi-hop tasks compared
to the next best solution. $\textit{Refiner}$ is a plug-and-play solution that
can be seamlessly integrated with RAG systems, facilitating its application
across diverse open-source frameworks. | cs.CL | Refiner: Restructure Retrieval Content Efficiently to Advance Question-Answering Capabilities |
2024-06-17T09:21:00Z | http://arxiv.org/pdf/2406.12937v1 | http://arxiv.org/abs/2406.12937v1 | Robert Flynn, Anton Ragni | 2024-06-17T09:21:00Z | eess.AS, cs.CL, cs.LG, cs.SD | When there is a mismatch between the training and test domains, current
speech recognition systems show significant performance degradation.
Self-training methods, such as noisy student teacher training, can help address
this and enable the adaptation of models under such domain shifts. However,
self-training typically requires a collection of unlabelled target domain data.
For settings where this is not practical, we investigate the benefit of
performing noisy student teacher training on recordings in the test set as a
test-time adaptation approach. Similarly to the dynamic evaluation approach in
language modelling, this enables the transfer of information across utterance
boundaries and functions as a method of domain adaptation. A range of in-domain
and out-of-domain datasets are used for experiments demonstrating large
relative gains of up to 32.2%. Interestingly, our method showed larger gains
than the typical self-training setup that utilises separate adaptation data. | eess.AS | Self-Train Before You Transcribe |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.