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
|
|---|---|---|---|---|---|---|---|---|
2024-06-04T23:36:08Z | http://arxiv.org/pdf/2406.02818v1 | http://arxiv.org/abs/2406.02818v1 | Yusen Zhang, Ruoxi Sun, Yanfei Chen, Tomas Pfister, Rui Zhang, Sercan Ö. Arik | 2024-06-04T23:36:08Z | cs.CL | Addressing the challenge of effectively processing long contexts has become a
critical issue for Large Language Models (LLMs). Two common strategies have
emerged: 1) reducing the input length, such as retrieving relevant chunks by
Retrieval-Augmented Generation (RAG), and 2) expanding the context window limit
of LLMs. However, both strategies have drawbacks: input reduction has no
guarantee of covering the part with needed information, while window extension
struggles with focusing on the pertinent information for solving the task. To
mitigate these limitations, we propose Chain-of-Agents (CoA), a novel framework
that harnesses multi-agent collaboration through natural language to enable
information aggregation and context reasoning across various LLMs over
long-context tasks. CoA consists of multiple worker agents who sequentially
communicate to handle different segmented portions of the text, followed by a
manager agent who synthesizes these contributions into a coherent final output.
CoA processes the entire input by interleaving reading and reasoning, and it
mitigates long context focus issues by assigning each agent a short context. We
perform comprehensive evaluation of CoA on a wide range of long-context tasks
in question answering, summarization, and code completion, demonstrating
significant improvements by up to 10% over strong baselines of RAG,
Full-Context, and multi-agent LLMs. | cs.CL | Chain of Agents: Large Language Models Collaborating on Long-Context Tasks |
2024-06-04T22:08:24Z | http://arxiv.org/pdf/2406.02804v1 | http://arxiv.org/abs/2406.02804v1 | François Roewer-Després, Jinyue Feng, Zining Zhu, Frank Rudzicz | 2024-06-04T22:08:24Z | cs.AI, cs.CL, cs.LG, I.2.0; I.2.7 | We present $\texttt{ACCORD}$, a framework and benchmark suite for
disentangling the commonsense grounding and reasoning abilities of large
language models (LLMs) through controlled, multi-hop counterfactuals.
$\texttt{ACCORD}$ introduces formal elements to commonsense reasoning to
explicitly control and quantify reasoning complexity beyond the typical 1 or 2
hops. Uniquely, $\texttt{ACCORD}$ can automatically generate benchmarks of
arbitrary reasoning complexity, and so it scales with future LLM improvements.
Benchmarking state-of-the-art LLMs -- including GPT-4o (2024-05-13),
Llama-3-70B-Instruct, and Mixtral-8x22B-Instruct-v0.1 -- shows performance
degrading to random chance with only moderate scaling, leaving substantial
headroom for improvement. We release a leaderboard of the benchmark suite
tested in this work, as well as code for automatically generating more complex
benchmarks. | cs.AI | $\texttt{ACCORD}$: Closing the Commonsense Measurability Gap |
2024-06-04T21:54:36Z | http://arxiv.org/pdf/2406.02798v2 | http://arxiv.org/abs/2406.02798v2 | Hao Peng, Huilian Sophie Qiu, Henrik Barslund Fosse, Brian Uzzi | 2024-06-07T20:56:35Z | cs.DL, cs.CL, cs.CY | How are the merits of innovative ideas communicated in science? Here we
conduct semantic analyses of grant application success with a focus on
scientific promotional language, which has been growing in frequency in many
contexts and purportedly may convey an innovative idea's originality and
significance. Our analysis attempts to surmount limitations of prior studies by
examining the full text of tens of thousands of both funded and unfunded grants
from three leading public and private funding agencies: the NIH, the NSF, and
the Novo Nordisk Foundation, one of the world's largest private science
foundations. We find a robust association between promotional language and the
support and adoption of innovative ideas by funders and other scientists.
First, the percentage of promotional language in a grant proposal is associated
with up to a doubling of the grant's probability of being funded. Second, a
grant's promotional language reflects its intrinsic level of innovativeness.
Third, the percentage of promotional language predicts the expected citation
and productivity impact of publications that are supported by funded grants.
Lastly, a computer-assisted experiment that manipulates the promotional
language in our data demonstrates how promotional language can communicate the
merit of ideas through cognitive activation. With the incidence of promotional
language in science steeply rising, and the pivotal role of grants in
converting promising and aspirational ideas into solutions, our analysis
provides empirical evidence that promotional language is associated with
effectively communicating the merits of innovative scientific ideas. | cs.DL | Promotional Language and the Adoption of Innovative Ideas in Science |
2024-06-04T21:43:56Z | http://arxiv.org/pdf/2406.02795v2 | http://arxiv.org/abs/2406.02795v2 | Priya Pitre, Kurt Luther | 2024-06-13T12:33:58Z | cs.HC, cs.CL | Opinion pieces (or op-eds) can provide valuable perspectives, but they often
represent only one side of a story, which can make readers susceptible to
confirmation bias and echo chambers. Exposure to different perspectives can
help readers overcome these obstacles and form more robust, nuanced views on
important societal issues. We designed ArguMentor, a human-AI collaboration
system that highlights claims in opinion pieces, identifies counter-arguments
for them using a LLM, and generates a context-based summary of based on current
events. It further enhances user understanding through additional features like
a Q&A bot (that answers user questions pertaining to the text), DebateMe (an
agent that users can argue any side of the piece with) and highlighting (where
users can highlight a word or passage to get its definition or context). Our
evaluation shows that participants can generate more arguments and
counter-arguments and have, on average, have more moderate views after engaging
with the system. | cs.HC | ArguMentor: Augmenting User Experiences with Counter-Perspectives |
2024-06-04T21:29:56Z | http://arxiv.org/pdf/2406.02791v1 | http://arxiv.org/abs/2406.02791v1 | Wang Zhu, Ishika Singh, Robin Jia, Jesse Thomason | 2024-06-04T21:29:56Z | cs.AI, cs.CL, cs.RO | Classical planning approaches guarantee finding a set of actions that can
achieve a given goal state when possible, but require an expert to specify
logical action semantics that govern the dynamics of the environment.
Researchers have shown that Large Language Models (LLMs) can be used to
directly infer planning steps based on commonsense knowledge and minimal domain
information alone, but such plans often fail on execution. We bring together
the strengths of classical planning and LLM commonsense inference to perform
domain induction, learning and validating action pre- and post-conditions based
on closed-loop interactions with the environment itself. We propose PSALM,
which leverages LLM inference to heuristically complete partial plans emitted
by a classical planner given partial domain knowledge, as well as to infer the
semantic rules of the domain in a logical language based on environment
feedback after execution. Our analysis on 7 environments shows that with just
one expert-curated example plans, using LLMs as heuristic planners and rule
predictors achieves lower environment execution steps and environment resets
than random exploration while simultaneously recovering the underlying ground
truth action semantics of the domain. | cs.AI | Language Models can Infer Action Semantics for Classical Planners from Environment Feedback |
2024-06-04T21:25:06Z | http://arxiv.org/pdf/2406.02787v1 | http://arxiv.org/abs/2406.02787v1 | Wenyue Hua, Kaijie Zhu, Lingyao Li, Lizhou Fan, Shuhang Lin, Mingyu Jin, Haochen Xue, Zelong Li, JinDong Wang, Yongfeng Zhang | 2024-06-04T21:25:06Z | cs.CL, cs.AI, cs.LG | This study intends to systematically disentangle pure logic reasoning and
text understanding by investigating the contrast across abstract and
contextualized logical problems from a comprehensive set of domains. We explore
whether LLMs demonstrate genuine reasoning capabilities across various domains
when the underlying logical structure remains constant. We focus on two main
questions (1) Can abstract logical problems alone accurately benchmark an LLM's
reasoning ability in real-world scenarios, disentangled from contextual support
in practical settings? (2) Does fine-tuning LLMs on abstract logic problem
generalize to contextualized logic problems and vice versa? To investigate
these questions, we focus on standard propositional logic, specifically
propositional deductive and abductive logic reasoning. In particular, we
construct instantiated datasets for deductive and abductive reasoning with 4
levels of difficulty, encompassing 12 distinct categories or domains based on
the categorization of Wikipedia. Our experiments aim to provide insights into
disentangling context in logical reasoning and the true reasoning capabilities
of LLMs and their generalization potential. The code and dataset are available
at: https://github.com/agiresearch/ContextHub. | cs.CL | Disentangling Logic: The Role of Context in Large Language Model Reasoning Capabilities |
2024-06-04T20:21:45Z | http://arxiv.org/pdf/2406.02756v1 | http://arxiv.org/abs/2406.02756v1 | Dehong Xu, Liang Qiu, Minseok Kim, Faisal Ladhak, Jaeyoung Do | 2024-06-04T20:21:45Z | cs.CL, cs.AI, cs.LG | Pre-trained large-scale language models (LLMs) excel at producing coherent
articles, yet their outputs may be untruthful, toxic, or fail to align with
user expectations. Current approaches focus on using reinforcement learning
with human feedback (RLHF) to improve model alignment, which works by
transforming coarse human preferences of LLM outputs into a feedback signal
that guides the model learning process. However, because this approach operates
on sequence-level feedback, it lacks the precision to identify the exact parts
of the output affecting user preferences. To address this gap, we propose a
method to enhance LLM alignment through fine-grained token-level supervision.
Specifically, we ask annotators to minimally edit less preferred responses
within the standard reward modeling dataset to make them more favorable,
ensuring changes are made only where necessary while retaining most of the
original content. The refined dataset is used to train a token-level reward
model, which is then used for training our fine-grained Proximal Policy
Optimization (PPO) model. Our experiment results demonstrate that this approach
can achieve up to an absolute improvement of $5.1\%$ in LLM performance, in
terms of win rate against the reference model, compared with the traditional
PPO model. | cs.CL | Aligning Large Language Models via Fine-grained Supervision |
2024-06-04T20:08:25Z | http://arxiv.org/pdf/2406.06582v2 | http://arxiv.org/abs/2406.06582v2 | Viet Anh Trinh, Rosy Southwell, Yiwen Guan, Xinlu He, Zhiyong Wang, Jacob Whitehill | 2024-06-25T17:44:00Z | cs.CL, cs.LG, eess.AS | Recent work on discrete speech tokenization has paved the way for models that
can seamlessly perform multiple tasks across modalities, e.g., speech
recognition, text to speech, speech to speech translation. Moreover, large
language models (LLMs) pretrained from vast text corpora contain rich
linguistic information that can improve accuracy in a variety of tasks. In this
paper, we present a decoder-only Discrete Multimodal Language Model (DMLM),
which can be flexibly applied to multiple tasks (ASR, T2S, S2TT, etc.) and
modalities (text, speech, vision). We explore several critical aspects of
discrete multi-modal models, including the loss function, weight
initialization, mixed training supervision, and codebook. Our results show that
DMLM benefits significantly, across multiple tasks and datasets, from a
combination of supervised and unsupervised training. Moreover, for ASR, it
benefits from initializing DMLM from a pretrained LLM, and from a codebook
derived from Whisper activations. | cs.CL | Discrete Multimodal Transformers with a Pretrained Large Language Model for Mixed-Supervision Speech Processing |
2024-06-04T20:02:52Z | http://arxiv.org/pdf/2406.02746v2 | http://arxiv.org/abs/2406.02746v2 | Jinghan Zhang, Xiting Wang, Weijieying Ren, Lu Jiang, Dongjie Wang, Kunpeng Liu | 2024-06-09T05:37:26Z | cs.CL | Large Language Models (LLMs) gain substantial reasoning and decision-making
capabilities from thought structures. However, existing methods such as Tree of
Thought and Retrieval Augmented Thoughts often fall short in complex tasks due
to the limitations of insufficient local retrieval of factual knowledge and
inadequate global selection of strategies. These limitations make it
challenging for these methods to balance factual accuracy and comprehensive
logical optimization effectively. To address these limitations, we introduce
the Retrieval Augmented Thought Tree (RATT), a novel thought structure that
considers both overall logical soundness and factual correctness at each step
of the thinking process. Specifically, at every point of a thought branch, RATT
performs planning and lookahead to explore and evaluate multiple potential
reasoning steps, and integrate the fact-checking ability of Retrieval-Augmented
Generation (RAG) with LLM's ability to assess overall strategy. Through this
combination of factual knowledge and strategic feasibility, the RATT adjusts
and integrates the thought tree structure to search for the most promising
branches within the search space. This thought structure significantly enhances
the model's coherence in logical inference and efficiency in decision-making,
and thus increases the limit of the capacity of LLM to generate reliable
inferences and decisions based on thought structures. A broad range of
experiments on different types of tasks showcases that the RATT structure
significantly outperforms existing methods in factual correctness and logical
coherence. | cs.CL | RATT: A Thought Structure for Coherent and Correct LLM Reasoning |
2024-06-04T19:22:13Z | http://arxiv.org/pdf/2406.02733v1 | http://arxiv.org/abs/2406.02733v1 | Min-Jae Hwang, Ilia Kulikov, Benjamin Peloquin, Hongyu Gong, Peng-Jen Chen, Ann Lee | 2024-06-04T19:22:13Z | cs.CL, cs.SD, eess.AS | In this paper, we propose a textless acoustic model with a self-supervised
distillation strategy for noise-robust expressive speech-to-speech translation
(S2ST). Recently proposed expressive S2ST systems have achieved impressive
expressivity preservation performances by cascading unit-to-speech (U2S)
generator to the speech-to-unit translation model. However, these systems are
vulnerable to the presence of noise in input speech, which is an assumption in
real-world translation scenarios. To address this limitation, we propose a U2S
generator that incorporates a distillation with no label (DINO) self-supervised
training strategy into it's pretraining process. Because the proposed method
captures noise-agnostic expressivity representation, it can generate qualified
speech even in noisy environment. Objective and subjective evaluation results
verified that the proposed method significantly improved the performance of the
expressive S2ST system in noisy environments while maintaining competitive
performance in clean environments. | cs.CL | Textless Acoustic Model with Self-Supervised Distillation for Noise-Robust Expressive Speech-to-Speech Translation |
2024-06-04T19:05:10Z | http://arxiv.org/pdf/2406.02721v2 | http://arxiv.org/abs/2406.02721v2 | Min Cai, Yuchen Zhang, Shichang Zhang, Fan Yin, Difan Zou, Yisong Yue, Ziniu Hu | 2024-06-18T15:58:38Z | cs.CL, cs.AI | We propose Self-Control, a novel method utilizing suffix gradients to control
the behavior of large language models (LLMs) without explicit human
annotations. Given a guideline expressed in suffix string and the model's
self-assessment of adherence, Self-Control computes the gradient of this
self-judgment concerning the model's hidden states, directly influencing the
auto-regressive generation process towards desired behaviors. To enhance
efficiency, we introduce Self-Control_{prefix}, a compact module that
encapsulates the learned representations from suffix gradients into a Prefix
Controller, facilitating inference-time control for various LLM behaviors. Our
experiments demonstrate Self-Control's efficacy across multiple domains,
including emotional modulation, ensuring harmlessness, and enhancing complex
reasoning. Especially, Self-Control_{prefix} enables a plug-and-play control
and jointly controls multiple attributes, improving model outputs without
altering model parameters or increasing inference-time costs. | cs.CL | Self-Control of LLM Behaviors by Compressing Suffix Gradient into Prefix Controller |
2024-06-04T17:58:18Z | http://arxiv.org/pdf/2406.02543v1 | http://arxiv.org/abs/2406.02543v1 | Yasin Abbasi Yadkori, Ilja Kuzborskij, András György, Csaba Szepesvári | 2024-06-04T17:58:18Z | cs.LG, cs.AI, cs.CL | We explore uncertainty quantification in large language models (LLMs), with
the goal to identify when uncertainty in responses given a query is large. We
simultaneously consider both epistemic and aleatoric uncertainties, where the
former comes from the lack of knowledge about the ground truth (such as about
facts or the language), and the latter comes from irreducible randomness (such
as multiple possible answers). In particular, we derive an
information-theoretic metric that allows to reliably detect when only epistemic
uncertainty is large, in which case the output of the model is unreliable. This
condition can be computed based solely on the output of the model obtained
simply by some special iterative prompting based on the previous responses.
Such quantification, for instance, allows to detect hallucinations (cases when
epistemic uncertainty is high) in both single- and multi-answer responses. This
is in contrast to many standard uncertainty quantification strategies (such as
thresholding the log-likelihood of a response) where hallucinations in the
multi-answer case cannot be detected. We conduct a series of experiments which
demonstrate the advantage of our formulation. Further, our investigations shed
some light on how the probabilities assigned to a given output by an LLM can be
amplified by iterative prompting, which might be of independent interest. | cs.LG | To Believe or Not to Believe Your LLM |
2024-06-04T17:56:28Z | http://arxiv.org/pdf/2406.02539v1 | http://arxiv.org/abs/2406.02539v1 | Hai-Long Sun, Da-Wei Zhou, Yang Li, Shiyin Lu, Chao Yi, Qing-Guo Chen, Zhao Xu, Weihua Luo, Kaifu Zhang, De-Chuan Zhan, Han-Jia Ye | 2024-06-04T17:56:28Z | cs.CV, cs.AI, cs.CL, cs.LG | The rapid development of Multimodal Large Language Models (MLLMs) like GPT-4V
has marked a significant step towards artificial general intelligence. Existing
methods mainly focus on aligning vision encoders with LLMs through supervised
fine-tuning (SFT) to endow LLMs with multimodal abilities, making MLLMs'
inherent ability to react to multiple languages progressively deteriorate as
the training process evolves. We empirically find that the imbalanced SFT
datasets, primarily composed of English-centric image-text pairs, lead to
significantly reduced performance in non-English languages. This is due to the
failure of aligning the vision encoder and LLM with multilingual tokens during
the SFT process. In this paper, we introduce Parrot, a novel method that
utilizes textual guidance to drive visual token alignment at the language
level. Parrot makes the visual tokens condition on diverse language inputs and
uses Mixture-of-Experts (MoE) to promote the alignment of multilingual tokens.
Specifically, to enhance non-English visual tokens alignment, we compute the
cross-attention using the initial visual features and textual embeddings, the
result of which is then fed into the MoE router to select the most relevant
experts. The selected experts subsequently convert the initial visual tokens
into language-specific visual tokens. Moreover, considering the current lack of
benchmarks for evaluating multilingual capabilities within the field, we
collect and make available a Massive Multilingual Multimodal Benchmark which
includes 6 languages, 15 categories, and 12,000 questions, named as MMMB. Our
method not only demonstrates state-of-the-art performance on multilingual
MMBench and MMMB, but also excels across a broad range of multimodal tasks.
Both the source code and the training dataset of Parrot will be made publicly
available. | cs.CV | Parrot: Multilingual Visual Instruction Tuning |
2024-06-04T17:55:43Z | http://arxiv.org/pdf/2406.02537v1 | http://arxiv.org/abs/2406.02537v1 | Chengzu Li, Caiqi Zhang, Han Zhou, Nigel Collier, Anna Korhonen, Ivan Vulić | 2024-06-04T17:55:43Z | cs.CL, cs.CV, cs.LG | Top-view perspective denotes a typical way in which humans read and reason
over different types of maps, and it is vital for localization and navigation
of humans as well as of `non-human' agents, such as the ones backed by large
Vision-Language Models (VLMs). Nonetheless, spatial reasoning capabilities of
modern VLMs remain unattested and underexplored. In this work, we thus study
their capability to understand and reason over spatial relations from the top
view. The focus on top view also enables controlled evaluations at different
granularity of spatial reasoning; we clearly disentangle different abilities
(e.g., recognizing particular objects versus understanding their relative
positions). We introduce the TopViewRS (Top-View Reasoning in Space) dataset,
consisting of 11,384 multiple-choice questions with either realistic or
semantic top-view map as visual input. We then use it to study and evaluate
VLMs across 4 perception and reasoning tasks with different levels of
complexity. Evaluation of 10 representative open- and closed-source VLMs
reveals the gap of more than 50% compared to average human performance, and it
is even lower than the random baseline in some cases. Although additional
experiments show that Chain-of-Thought reasoning can boost model capabilities
by 5.82% on average, the overall performance of VLMs remains limited. Our
findings underscore the critical need for enhanced model capability in top-view
spatial reasoning and set a foundation for further research towards human-level
proficiency of VLMs in real-world multimodal tasks. | cs.CL | TopViewRS: Vision-Language Models as Top-View Spatial Reasoners |
2024-06-04T17:55:38Z | http://arxiv.org/pdf/2406.02536v1 | http://arxiv.org/abs/2406.02536v1 | Yijiong Yu, Huiqiang Jiang, Xufang Luo, Qianhui Wu, Chin-Yew Lin, Dongsheng Li, Yuqing Yang, Yongfeng Huang, Lili Qiu | 2024-06-04T17:55:38Z | cs.CL, cs.LG | Large Language Models (LLMs) are increasingly applied in various real-world
scenarios due to their excellent generalization capabilities and robust
generative abilities. However, they exhibit position bias, also known as "lost
in the middle", a phenomenon that is especially pronounced in long-context
scenarios, which indicates the placement of the key information in different
positions of a prompt can significantly affect accuracy. This paper first
explores the micro-level manifestations of position bias, concluding that
attention weights are a micro-level expression of position bias. It further
identifies that, in addition to position embeddings, causal attention mask also
contributes to position bias by creating position-specific hidden states. Based
on these insights, we propose a method to mitigate position bias by scaling
this positional hidden states. Experiments on the NaturalQuestions
Multi-document QA, KV retrieval, LongBench and timeline reorder tasks, using
various models including RoPE models, context windowextended models, and Alibi
models, demonstrate the effectiveness and generalizability of our approach. Our
method can improve performance by up to 15.2% by modifying just one dimension
of hidden states. Our code is available at https://aka.ms/PositionalHidden. | cs.CL | Mitigate Position Bias in Large Language Models via Scaling a Single Dimension |
2024-06-04T17:53:36Z | http://arxiv.org/pdf/2406.02532v2 | http://arxiv.org/abs/2406.02532v2 | Ruslan Svirschevski, Avner May, Zhuoming Chen, Beidi Chen, Zhihao Jia, Max Ryabinin | 2024-06-25T19:35:06Z | cs.CL | As large language models gain widespread adoption, running them efficiently
becomes crucial. Recent works on LLM inference use speculative decoding to
achieve extreme speedups. However, most of these works implicitly design their
algorithms for high-end datacenter hardware. In this work, we ask the opposite
question: how fast can we run LLMs on consumer machines? Consumer GPUs can no
longer fit the largest available models (50B+ parameters) and must offload them
to RAM or SSD. When running with offloaded parameters, the inference engine can
process batches of hundreds or thousands of tokens at the same time as just one
token, making it a natural fit for speculative decoding. We propose SpecExec
(Speculative Execution), a simple parallel decoding method that can generate up
to 20 tokens per target model iteration for popular LLM families. It utilizes
the high spikiness of the token probabilities distribution in modern LLMs and a
high degree of alignment between model output probabilities. SpecExec takes the
most probable tokens continuation from the draft model to build a "cache" tree
for the target model, which then gets validated in a single pass. Using
SpecExec, we demonstrate inference of 50B+ parameter LLMs on consumer GPUs with
RAM offloading at 4-6 tokens per second with 4-bit quantization or 2-3 tokens
per second with 16-bit weights. | cs.CL | SpecExec: Massively Parallel Speculative Decoding for Interactive LLM Inference on Consumer Devices |
2024-06-04T17:50:34Z | http://arxiv.org/pdf/2406.02528v5 | http://arxiv.org/abs/2406.02528v5 | Rui-Jie Zhu, Yu Zhang, Ethan Sifferman, Tyler Sheaves, Yiqiao Wang, Dustin Richmond, Peng Zhou, Jason K. Eshraghian | 2024-06-18T17:30:06Z | cs.CL | Matrix multiplication (MatMul) typically dominates the overall computational
cost of large language models (LLMs). This cost only grows as LLMs scale to
larger embedding dimensions and context lengths. In this work, we show that
MatMul operations can be completely eliminated from LLMs while maintaining
strong performance at billion-parameter scales. Our experiments show that our
proposed MatMul-free models achieve performance on-par with state-of-the-art
Transformers that require far more memory during inference at a scale up to at
least 2.7B parameters. We investigate the scaling laws and find that the
performance gap between our MatMul-free models and full precision Transformers
narrows as the model size increases. We also provide a GPU-efficient
implementation of this model which reduces memory usage by up to 61% over an
unoptimized baseline during training. By utilizing an optimized kernel during
inference, our model's memory consumption can be reduced by more than 10x
compared to unoptimized models. To properly quantify the efficiency of our
architecture, we build a custom hardware solution on an FPGA which exploits
lightweight operations beyond what GPUs are capable of. We processed
billion-parameter scale models at 13W beyond human readable throughput, moving
LLMs closer to brain-like efficiency. This work not only shows how far LLMs can
be stripped back while still performing effectively, but also points at the
types of operations future accelerators should be optimized for in processing
the next generation of lightweight LLMs. Our code implementation is available
at https://github.com/ridgerchu/matmulfreellm. | cs.CL | Scalable MatMul-free Language Modeling |
2024-06-04T17:45:26Z | http://arxiv.org/pdf/2406.02657v1 | http://arxiv.org/abs/2406.02657v1 | Namgyu Ho, Sangmin Bae, Taehyeon Kim, Hyunjik Jo, Yireun Kim, Tal Schuster, Adam Fisch, James Thorne, Se-Young Yun | 2024-06-04T17:45:26Z | cs.CL, cs.AI, cs.LG | This paper presents the Block Transformer architecture which adopts
hierarchical global-to-local modeling to autoregressive transformers to
mitigate the inference bottlenecks of self-attention. To apply self-attention,
the key-value (KV) cache of all previous sequences must be retrieved from
memory at every decoding step. Thereby, this KV cache IO becomes a significant
bottleneck in batch inference. We notice that these costs stem from applying
self-attention on the global context, therefore we isolate the expensive
bottlenecks of global modeling to lower layers and apply fast local modeling in
upper layers. To mitigate the remaining costs in the lower layers, we aggregate
input tokens into fixed size blocks and then apply self-attention at this
coarse level. Context information is aggregated into a single embedding to
enable upper layers to decode the next block of tokens, without global
attention. Free of global attention bottlenecks, the upper layers can fully
utilize the compute hardware to maximize inference throughput. By leveraging
global and local modules, the Block Transformer architecture demonstrates
10-20x gains in inference throughput compared to vanilla transformers with
equivalent perplexity. Our work introduces a new approach to optimize language
model inference through novel application of global-to-local modeling. Code is
available at https://github.com/itsnamgyu/block-transformer. | cs.CL | Block Transformer: Global-to-Local Language Modeling for Fast Inference |
2024-06-04T17:42:21Z | http://arxiv.org/pdf/2406.02524v2 | http://arxiv.org/abs/2406.02524v2 | Maciej Besta, Lorenzo Paleari, Ales Kubicek, Piotr Nyczyk, Robert Gerstenberger, Patrick Iff, Tomasz Lehmann, Hubert Niewiadomski, Torsten Hoefler | 2024-06-07T17:58:22Z | cs.CL | Large Language Models (LLMs) are revolutionizing various domains, yet
verifying their answers remains a significant challenge, especially for
intricate open-ended tasks such as consolidation, summarization, and extraction
of knowledge. In this work, we propose CheckEmbed: an accurate, scalable, and
simple LLM verification approach. CheckEmbed is driven by a straightforward yet
powerful idea: in order to compare LLM solutions to one another or to the
ground-truth, compare their corresponding answer-level embeddings obtained with
a model such as GPT Text Embedding Large. This reduces a complex textual answer
to a single embedding, facilitating straightforward, fast, and meaningful
verification. We develop a comprehensive verification pipeline implementing the
CheckEmbed methodology. The CheckEmbed pipeline also comes with metrics for
assessing the truthfulness of the LLM answers, such as embedding heatmaps and
their summaries. We show how to use these metrics for deploying practical
engines that decide whether an LLM answer is satisfactory or not. We apply the
pipeline to real-world document analysis tasks, including term extraction and
document summarization, showcasing significant improvements in accuracy,
cost-effectiveness, and runtime performance compared to existing token-,
sentence-, and fact-level schemes such as BERTScore or SelfCheckGPT. | cs.CL | CheckEmbed: Effective Verification of LLM Solutions to Open-Ended Tasks |
2024-06-04T17:39:23Z | http://arxiv.org/pdf/2406.02517v1 | http://arxiv.org/abs/2406.02517v1 | Jiashu Yao, Heyan Huang, Zeming Liu, Yuhang Guo | 2024-06-04T17:39:23Z | cs.CL | Data augmentation is an effective way to diversify corpora in machine
translation, but previous methods may introduce semantic inconsistency between
original and augmented data because of irreversible operations and random
subword sampling procedures. To generate both symbolically diverse and
semantically consistent augmentation data, we propose Deterministic Reversible
Data Augmentation (DRDA), a simple but effective data augmentation method for
neural machine translation. DRDA adopts deterministic segmentations and
reversible operations to generate multi-granularity subword representations and
pulls them closer together with multi-view techniques. With no extra corpora or
model changes required, DRDA outperforms strong baselines on several
translation tasks with a clear margin (up to 4.3 BLEU gain over Transformer)
and exhibits good robustness in noisy, low-resource, and cross-domain datasets. | cs.CL | Deterministic Reversible Data Augmentation for Neural Machine Translation |
2024-06-04T16:59:11Z | http://arxiv.org/pdf/2406.02488v1 | http://arxiv.org/abs/2406.02488v1 | Hao Yen, Pin-Jui Ku, Sabato Marco Siniscalchi, Chin-Hui Lee | 2024-06-04T16:59:11Z | eess.AS, cs.CL, cs.SD | We propose a novel language-universal approach to end-to-end automatic spoken
keyword recognition (SKR) leveraging upon (i) a self-supervised pre-trained
model, and (ii) a set of universal speech attributes (manner and place of
articulation). Specifically, Wav2Vec2.0 is used to generate robust speech
representations, followed by a linear output layer to produce attribute
sequences. A non-trainable pronunciation model then maps sequences of
attributes into spoken keywords in a multilingual setting. Experiments on the
Multilingual Spoken Words Corpus show comparable performances to character- and
phoneme-based SKR in seen languages. The inclusion of domain adversarial
training (DAT) improves the proposed framework, outperforming both character-
and phoneme-based SKR approaches with 13.73% and 17.22% relative word error
rate (WER) reduction in seen languages, and achieves 32.14% and 19.92% WER
reduction for unseen languages in zero-shot settings. | eess.AS | Language-Universal Speech Attributes Modeling for Zero-Shot Multilingual Spoken Keyword Recognition |
2024-06-04T16:49:06Z | http://arxiv.org/pdf/2406.02481v1 | http://arxiv.org/abs/2406.02481v1 | Jakub Hoscilowicz, Pawel Popiolek, Jan Rudkowski, Jedrzej Bieniasz, Artur Janicki | 2024-06-04T16:49:06Z | cs.CL, cs.CR | With the help of simple fine-tuning, one can artificially embed hidden text
into large language models (LLMs). This text is revealed only when triggered by
a specific query to the LLM. Two primary applications are LLM fingerprinting
and steganography. In the context of LLM fingerprinting, a unique text
identifier (fingerprint) is embedded within the model to verify licensing
compliance. In the context of steganography, the LLM serves as a carrier for
hidden messages that can be disclosed through a designated trigger.
Our work demonstrates that embedding hidden text in the LLM via fine-tuning,
though seemingly secure due to the vast number of potential triggers (any
sequence of characters or tokens could serve as a trigger), is susceptible to
extraction through analysis of the LLM's output decoding process. We propose a
novel approach to extraction called Unconditional Token Forcing. It is premised
on the hypothesis that iteratively feeding each token from the LLM's vocabulary
into the model should reveal sequences with abnormally high token
probabilities, indicating potential embedded text candidates. Additionally, our
experiments show that when the first token of a hidden fingerprint is used as
an input, the LLM not only produces an output sequence with high token
probabilities, but also repetitively generates the fingerprint itself. We also
present a method to hide text in such a way that it is resistant to
Unconditional Token Forcing, which we named Unconditional Token Forcing
Confusion. | cs.CL | Hiding Text in Large Language Models: Introducing Unconditional Token Forcing Confusion |
2024-06-04T16:42:17Z | http://arxiv.org/pdf/2406.02472v1 | http://arxiv.org/abs/2406.02472v1 | Zhihan Zhang, Yixin Cao, Chenchen Ye, Yunshan Ma, Lizi Liao, Tat-Seng Chua | 2024-06-04T16:42:17Z | cs.CL | The digital landscape is rapidly evolving with an ever-increasing volume of
online news, emphasizing the need for swift and precise analysis of complex
events. We refer to the complex events composed of many news articles over an
extended period as Temporal Complex Event (TCE). This paper proposes a novel
approach using Large Language Models (LLMs) to systematically extract and
analyze the event chain within TCE, characterized by their key points and
timestamps. We establish a benchmark, named TCELongBench, to evaluate the
proficiency of LLMs in handling temporal dynamics and understanding extensive
text. This benchmark encompasses three distinct tasks - reading comprehension,
temporal sequencing, and future event forecasting. In the experiment, we
leverage retrieval-augmented generation (RAG) method and LLMs with long context
window to deal with lengthy news articles of TCE. Our findings indicate that
models with suitable retrievers exhibit comparable performance with those
utilizing long context window. | cs.CL | Analyzing Temporal Complex Events with Large Language Models? A Benchmark towards Temporal, Long Context Understanding |
2024-06-04T16:38:57Z | http://arxiv.org/pdf/2406.02469v1 | http://arxiv.org/abs/2406.02469v1 | Stefani Karp, Nikunj Saunshi, Sobhan Miryoosefi, Sashank J. Reddi, Sanjiv Kumar | 2024-06-04T16:38:57Z | cs.LG, cs.CL | Recently, there has been increasing interest in efficient pretraining
paradigms for training Transformer-based models. Several recent approaches use
smaller models to initialize larger models in order to save computation (e.g.,
stacking and fusion). In this work, we study the fundamental question of how to
select the best growing strategy from a given pool of growing strategies. Prior
works have extensively focused on loss- and/or function-preserving behavior at
initialization or simply performance at the end of training. Instead, we
identify that behavior at initialization can be misleading as a predictor of
final performance and present an alternative perspective based on early
training dynamics, which we call "landscape-aware growing (LAG)". We perform
extensive analysis of correlation of the final performance with performance in
the initial steps of training and find early and more accurate predictions of
the optimal growing strategy (i.e., with only a small "lag" after
initialization). This perspective also motivates an adaptive strategy for
gradual stacking. | cs.LG | Landscape-Aware Growing: The Power of a Little LAG |
2024-06-04T16:14:00Z | http://arxiv.org/pdf/2406.02449v1 | http://arxiv.org/abs/2406.02449v1 | Henry Conklin, Kenny Smith | 2024-06-04T16:14:00Z | cs.CL, cs.AI | Large scale neural models show impressive performance across a wide array of
linguistic tasks. Despite this they remain, largely, black-boxes - inducing
vector-representations of their input that prove difficult to interpret. This
limits our ability to understand what they learn, and when the learn it, or
describe what kinds of representations generalise well out of distribution. To
address this we introduce a novel approach to interpretability that looks at
the mapping a model learns from sentences to representations as a kind of
language in its own right. In doing so we introduce a set of
information-theoretic measures that quantify how structured a model's
representations are with respect to its input, and when during training that
structure arises. Our measures are fast to compute, grounded in linguistic
theory, and can predict which models will generalise best based on their
representations. We use these measures to describe two distinct phases of
training a transformer: an initial phase of in-distribution learning which
reduces task loss, then a second stage where representations becoming robust to
noise. Generalisation performance begins to increase during this second phase,
drawing a link between generalisation and robustness to noise. Finally we look
at how model size affects the structure of the representational space, showing
that larger models ultimately compress their representations more than their
smaller counterparts. | cs.CL | Representations as Language: An Information-Theoretic Framework for Interpretability |
2024-06-04T16:09:13Z | http://arxiv.org/pdf/2406.06581v2 | http://arxiv.org/abs/2406.06581v2 | Reid McIlroy-Young, Katrina Brown, Conlan Olson, Linjun Zhang, Cynthia Dwork | 2024-06-12T13:59:13Z | cs.CL, cs.AI, cs.LG | The development of generative language models that can create long and
coherent textual outputs via autoregression has lead to a proliferation of uses
and a corresponding sweep of analyses as researches work to determine the
limitations of this new paradigm. Unlike humans, these 'Large Language Models'
(LLMs) are highly sensitive to small changes in their inputs, leading to
unwanted inconsistency in their behavior. One problematic inconsistency when
LLMs are used to answer multiple-choice questions or analyze multiple inputs is
order dependency: the output of an LLM can (and often does) change
significantly when sub-sequences are swapped, despite both orderings being
semantically identical. In this paper we present Set-Based Prompting, a
technique that guarantees the output of an LLM will not have order dependence
on a specified set of sub-sequences. We show that this method provably
eliminates order dependency, and that it can be applied to any
transformer-based LLM to enable text generation that is unaffected by
re-orderings. Delving into the implications of our method, we show that,
despite our inputs being out of distribution, the impact on expected accuracy
is small, where the expectation is over the order of uniformly chosen shuffling
of the candidate responses, and usually significantly less in practice. Thus,
Set-Based Prompting can be used as a 'dropped-in' method on fully trained
models. Finally, we discuss how our method's success suggests that other strong
guarantees can be obtained on LLM performance via modifying the input
representations. | cs.CL | Set-Based Prompting: Provably Solving the Language Model Order Dependency Problem |
2024-06-04T15:11:27Z | http://arxiv.org/pdf/2406.02396v1 | http://arxiv.org/abs/2406.02396v1 | Kenneth Enevoldsen, Márton Kardos, Niklas Muennighoff, Kristoffer Laigaard Nielbo | 2024-06-04T15:11:27Z | cs.CL, cs.AI | The evaluation of English text embeddings has transitioned from evaluating a
handful of datasets to broad coverage across many tasks through benchmarks such
as MTEB. However, this is not the case for multilingual text embeddings due to
a lack of available benchmarks. To address this problem, we introduce the
Scandinavian Embedding Benchmark (SEB). SEB is a comprehensive framework that
enables text embedding evaluation for Scandinavian languages across 24 tasks,
10 subtasks, and 4 task categories. Building on SEB, we evaluate more than 26
models, uncovering significant performance disparities between public and
commercial solutions not previously captured by MTEB. We open-source SEB and
integrate it with MTEB, thus bridging the text embedding evaluation gap for
Scandinavian languages. | cs.CL | The Scandinavian Embedding Benchmarks: Comprehensive Assessment of Multilingual and Monolingual Text Embedding |
2024-06-04T15:08:56Z | http://arxiv.org/pdf/2406.02394v1 | http://arxiv.org/abs/2406.02394v1 | Maxime Griot, Jean Vanderdonckt, Demet Yuksel, Coralie Hemptinne | 2024-06-04T15:08:56Z | cs.CL, cs.AI, cs.LG | Large Language Models (LLMs) like ChatGPT demonstrate significant potential
in the medical field, often evaluated using multiple-choice questions (MCQs)
similar to those found on the USMLE. Despite their prevalence in medical
education, MCQs have limitations that might be exacerbated when assessing LLMs.
To evaluate the effectiveness of MCQs in assessing the performance of LLMs, we
developed a fictional medical benchmark focused on a non-existent gland, the
Glianorex. This approach allowed us to isolate the knowledge of the LLM from
its test-taking abilities. We used GPT-4 to generate a comprehensive textbook
on the Glianorex in both English and French and developed corresponding
multiple-choice questions in both languages. We evaluated various open-source,
proprietary, and domain-specific LLMs using these questions in a zero-shot
setting. The models achieved average scores around 67%, with minor performance
differences between larger and smaller models. Performance was slightly higher
in English than in French. Fine-tuned medical models showed some improvement
over their base versions in English but not in French. The uniformly high
performance across models suggests that traditional MCQ-based benchmarks may
not accurately measure LLMs' clinical knowledge and reasoning abilities,
instead highlighting their pattern recognition skills. This study underscores
the need for more robust evaluation methods to better assess the true
capabilities of LLMs in medical contexts. | cs.CL | Multiple Choice Questions and Large Languages Models: A Case Study with Fictional Medical Data |
2024-06-04T14:55:43Z | http://arxiv.org/pdf/2406.02378v1 | http://arxiv.org/abs/2406.02378v1 | Guangliang Liu, Haitao Mao, Bochuan Cao, Zhiyu Xue, Kristen Johnson, Jiliang Tang, Rongrong Wang | 2024-06-04T14:55:43Z | cs.CL | Large Language Models (LLMs) can improve their responses when instructed to
do so, a capability known as self-correction. When these instructions lack
specific details about the issues in the response, this is referred to as
leveraging the intrinsic self-correction capability. The empirical success of
self-correction can be found in various applications, e.g., text detoxification
and social bias mitigation. However, leveraging this self-correction capability
may not always be effective, as it has the potential to revise an initially
correct response into an incorrect one. In this paper, we endeavor to
understand how and why leveraging the self-correction capability is effective.
We identify that appropriate instructions can guide LLMs to a convergence
state, wherein additional self-correction steps do not yield further
performance improvements. We empirically demonstrate that model uncertainty and
activated latent concepts jointly characterize the effectiveness of
self-correction. Furthermore, we provide a mathematical formulation indicating
that the activated latent concept drives the convergence of the model
uncertainty and self-correction performance. Our analysis can also be
generalized to the self-correction behaviors observed in Vision-Language Models
(VLMs). Moreover, we highlight that task-agnostic debiasing can benefit from
our principle in terms of selecting effective fine-tuning samples. Such initial
success demonstrates the potential extensibility for better instruction tuning
and safety alignment. | cs.CL | On the Intrinsic Self-Correction Capability of LLMs: Uncertainty and Latent Concept |
2024-06-04T14:55:14Z | http://arxiv.org/pdf/2406.02377v1 | http://arxiv.org/abs/2406.02377v1 | Qiyao Ma, Xubin Ren, Chao Huang | 2024-06-04T14:55:14Z | cs.IR, cs.AI, cs.CL | Recommender systems help users navigate information overload by providing
personalized recommendations aligned with their preferences. Collaborative
Filtering (CF) is a widely adopted approach, but while advanced techniques like
graph neural networks (GNNs) and self-supervised learning (SSL) have enhanced
CF models for better user representations, they often lack the ability to
provide explanations for the recommended items. Explainable recommendations aim
to address this gap by offering transparency and insights into the
recommendation decision-making process, enhancing users' understanding. This
work leverages the language capabilities of Large Language Models (LLMs) to
push the boundaries of explainable recommender systems. We introduce a
model-agnostic framework called XRec, which enables LLMs to provide
comprehensive explanations for user behaviors in recommender systems. By
integrating collaborative signals and designing a lightweight collaborative
adaptor, the framework empowers LLMs to understand complex patterns in
user-item interactions and gain a deeper understanding of user preferences. Our
extensive experiments demonstrate the effectiveness of XRec, showcasing its
ability to generate comprehensive and meaningful explanations that outperform
baseline approaches in explainable recommender systems. We open-source our
model implementation at https://github.com/HKUDS/XRec. | cs.IR | XRec: Large Language Models for Explainable Recommendation |
2024-06-04T14:53:24Z | http://arxiv.org/pdf/2406.02376v2 | http://arxiv.org/abs/2406.02376v2 | Zhiwei Cao, Qian Cao, Yu Lu, Ningxin Peng, Luyang Huang, Shanbo Cheng, Jinsong Su | 2024-06-17T15:02:11Z | cs.CL | The growing popularity of Large Language Models has sparked interest in
context compression for Large Language Models (LLMs). However, the performance
of previous methods degrades dramatically as compression ratios increase,
sometimes even falling to the closed-book level. This decline can be attributed
to the loss of key information during the compression process. Our preliminary
study supports this hypothesis, emphasizing the significance of retaining key
information to maintain model performance under high compression ratios. As a
result, we introduce Query-Guided Compressor (QGC), which leverages queries to
guide the context compression process, effectively preserving key information
within the compressed context. Additionally, we employ a dynamic compression
strategy. We validate the effectiveness of our proposed QGC on the Question
Answering task, including NaturalQuestions, TriviaQA, and HotpotQA datasets.
Experimental results show that QGC can consistently perform well even at high
compression ratios, which also offers significant benefits in terms of
inference cost and throughput. | cs.CL | Retaining Key Information under High Compression Ratios: Query-Guided Compressor for LLMs |
2024-06-04T14:46:25Z | http://arxiv.org/pdf/2406.02368v1 | http://arxiv.org/abs/2406.02368v1 | Jianghao Lin, Xinyi Dai, Rong Shan, Bo Chen, Ruiming Tang, Yong Yu, Weinan Zhang | 2024-06-04T14:46:25Z | cs.IR, cs.CL | Large language models (LLMs) have achieved remarkable progress in the field
of natural language processing (NLP), demonstrating remarkable abilities in
producing text that resembles human language for various tasks. This opens up
new opportunities for employing them in recommender systems (RSs). In this
paper, we specifically examine the sample efficiency of LLM-enhanced
recommender systems, which pertains to the model's capacity to attain superior
performance with a limited quantity of training data. Conventional
recommendation models (CRMs) often need a large amount of training data because
of the sparsity of features and interactions. Hence, we propose and verify our
core viewpoint: Large Language Models Make Sample-Efficient Recommender
Systems. We propose a simple yet effective framework (i.e., Laser) to validate
the viewpoint from two aspects: (1) LLMs themselves are sample-efficient
recommenders; and (2) LLMs, as feature generators and encoders, make CRMs more
sample-efficient. Extensive experiments on two public datasets show that Laser
requires only a small fraction of training samples to match or even surpass
CRMs that are trained on the entire training set, demonstrating superior sample
efficiency. | cs.IR | Large Language Models Make Sample-Efficient Recommender Systems |
2024-06-04T14:34:39Z | http://arxiv.org/pdf/2406.02356v1 | http://arxiv.org/abs/2406.02356v1 | Andrew Gambardella, Yusuke Iwasawa, Yutaka Matsuo | 2024-06-04T14:34:39Z | cs.LG, cs.AI, cs.CL | The ability (and inability) of large language models (LLMs) to perform
arithmetic tasks has been the subject of much theoretical and practical debate.
We show that LLMs are frequently able to correctly and confidently predict the
first digit of n-digit by m-digit multiplication tasks without using chain of
thought reasoning, despite these tasks require compounding operations to solve.
Simultaneously, LLMs in practice often fail to correctly or confidently predict
the last digit of an n-digit by m-digit multiplication, a task equivalent to
1-digit by 1-digit multiplication which can be easily learned or memorized. We
show that the latter task can be solved more robustly when the LLM is
conditioned on all of the correct higher-order digits, which on average
increases the confidence of the correct last digit on 5-digit by 5-digit
multiplication tasks using Llama 2-13B by over 230% (0.13 to 0.43) and
Mistral-7B by 150% (0.22 to 0.55). | cs.LG | Language Models Do Hard Arithmetic Tasks Easily and Hardly Do Easy Arithmetic Tasks |
2024-06-04T14:24:53Z | http://arxiv.org/pdf/2406.02350v2 | http://arxiv.org/abs/2406.02350v2 | Maojun Sun | 2024-06-05T15:08:42Z | cs.CL, cs.AI | Large language models (LLMs) have shown amazing capabilities in knowledge
memorization and the present. However, when it comes to domain-specific
knowledge and downstream tasks like medical, general LLMs are often unable to
give precise answers. In addition, when people want LLMs to answer
classification questions, they usually go through instruction tuning first.
However, LLMs do not always give a direct index of the categorization after
instruction tuning. In this paper, we proposed LlamaCare, a fine-tuned medical
language model, and Extended Classification Integration(ECI), a module to
handle classification problems of LLMs. Our contributions are : (i) We
fine-tuned a large language model of medical knowledge with very low carbon
emissions and achieved similar performance with ChatGPT by a 24G GPU. (ii) We
solved the problem of redundant categorical answers and improved the
performance of LLMs by proposing a new module called Extended Classification
Integration. (iii) We released our processed data for one-shot and few-shot
training for some benchmarks such as PubMedQA and USMLE 1-3 step. Our method
achieves a close performance comparable to some state-of-the-art models with
the same quantity of parameters on benchmarks, while being more environmentally
friendly by using less GPU computation time. Our models, codes, and datasets
can be found at \url{https://github.com/Stephen-SMJ/LLamaCare}. | cs.CL | LlamaCare: A Large Medical Language Model for Enhancing Healthcare Knowledge Sharing |
2024-06-04T14:09:36Z | http://arxiv.org/pdf/2406.02338v1 | http://arxiv.org/abs/2406.02338v1 | Michele Mastromattei, Fabio Massimo Zanzotto | 2024-06-04T14:09:36Z | cs.CL, cs.AI | This paper explores the correlation between linguistic diversity, sentiment
analysis and transformer model architectures. We aim to investigate how
different English variations impact transformer-based models for irony
detection. To conduct our study, we used the EPIC corpus to extract five
diverse English variation-specific datasets and applied the KEN pruning
algorithm on five different architectures. Our results reveal several
similarities between optimal subnetworks, which provide insights into the
linguistic variations that share strong resemblances and those that exhibit
greater dissimilarities. We discovered that optimal subnetworks across models
share at least 60% of their parameters, emphasizing the significance of
parameter values in capturing and interpreting linguistic variations. This
study highlights the inherent structural similarities between models trained on
different variants of the same language and also the critical role of parameter
values in capturing these nuances. | cs.CL | Linguistic Fingerprint in Transformer Models: How Language Variation Influences Parameter Selection in Irony Detection |
2024-06-04T14:06:03Z | http://arxiv.org/pdf/2406.02335v1 | http://arxiv.org/abs/2406.02335v1 | Anisia Katinskaia, Roman Yangarber | 2024-06-04T14:06:03Z | cs.CL | We investigate how pretrained language models (PLM) encode the grammatical
category of verbal aspect in Russian. Encoding of aspect in transformer LMs has
not been studied previously in any language. A particular challenge is posed by
"alternative contexts": where either the perfective or the imperfective aspect
is suitable grammatically and semantically. We perform probing using BERT and
RoBERTa on alternative and non-alternative contexts. First, we assess the
models' performance on aspect prediction, via behavioral probing. Next, we
examine the models' performance when their contextual representations are
substituted with counterfactual representations, via causal probing. These
counterfactuals alter the value of the "boundedness" feature--a semantic
feature, which characterizes the action in the context. Experiments show that
BERT and RoBERTa do encode aspect--mostly in their final layers. The
counterfactual interventions affect perfective and imperfective in opposite
ways, which is consistent with grammar: perfective is positively affected by
adding the meaning of boundedness, and vice versa. The practical implications
of our probing results are that fine-tuning only the last layers of BERT on
predicting aspect is faster and more effective than fine-tuning the whole
model. The model has high predictive uncertainty about aspect in alternative
contexts, which tend to lack explicit hints about the boundedness of the
described action. | cs.CL | Probing the Category of Verbal Aspect in Transformer Language Models |
2024-06-04T14:02:53Z | http://arxiv.org/pdf/2406.06580v1 | http://arxiv.org/abs/2406.06580v1 | Mengru Ding, Hanmeng Liu, Zhizhang Fu, Jian Song, Wenbo Xie, Yue Zhang | 2024-06-04T14:02:53Z | cs.CL, cs.AI | Recent advancements in Chain-of-Thought (CoT) reasoning utilize complex
modules but are hampered by high token consumption, limited applicability, and
challenges in reproducibility. This paper conducts a critical evaluation of CoT
prompting, extending beyond arithmetic to include complex logical and
commonsense reasoning tasks, areas where standard CoT methods fall short. We
propose the integration of human-like heuristics and shortcuts into language
models (LMs) through "break the chain" strategies. These strategies disrupt
traditional CoT processes using controlled variables to assess their efficacy.
Additionally, we develop innovative zero-shot prompting strategies that
encourage the use of shortcuts, enabling LMs to quickly exploit reasoning clues
and bypass detailed procedural steps. Our comprehensive experiments across
various LMs, both commercial and open-source, reveal that LMs maintain
effective performance with "break the chain" strategies. We also introduce
ShortcutQA, a dataset specifically designed to evaluate reasoning through
shortcuts, compiled from competitive tests optimized for heuristic reasoning
tasks such as forward/backward reasoning and simplification. Our analysis
confirms that ShortcutQA not only poses a robust challenge to LMs but also
serves as an essential benchmark for enhancing reasoning efficiency in AI. | cs.CL | Break the Chain: Large Language Models Can be Shortcut Reasoners |
2024-06-04T14:00:25Z | http://arxiv.org/pdf/2406.02332v1 | http://arxiv.org/abs/2406.02332v1 | Phoebe Klett, Thomas Ahle | 2024-06-04T14:00:25Z | cs.LG, cs.CL | Pre-trained language models demonstrate general intelligence and common
sense, but long inputs quickly become a bottleneck for memorizing information
at inference time. We resurface a simple method, Memorizing Transformers (Wu et
al., 2022), that gives the model access to a bank of pre-computed memories. We
show that it is possible to fix many of the shortcomings of the original
method, such as the need for fine-tuning, by critically assessing how
positional encodings should be updated for the keys and values retrieved. This
intuitive method uses the model's own key/query system to select and attend to
the most relevant memories at each generation step, rather than using external
embeddings. We demonstrate the importance of external information being
retrieved in a majority of decoder layers, contrary to previous work. We open
source a new counterfactual long-range retrieval benchmark, and show that
Extended Mind Transformers outperform today's state of the art by 6% on
average. | cs.LG | Extended Mind Transformers |
2024-06-04T14:00:02Z | http://arxiv.org/pdf/2406.02331v1 | http://arxiv.org/abs/2406.02331v1 | ChaeHun Park, Koanho Lee, Hyesu Lim, Jaeseok Kim, Junmo Park, Yu-Jung Heo, Du-Seong Chang, Jaegul Choo | 2024-06-04T14:00:02Z | cs.CL | Building a reliable visual question answering~(VQA) system across different
languages is a challenging problem, primarily due to the lack of abundant
samples for training. To address this challenge, recent studies have employed
machine translation systems for the cross-lingual VQA task. This involves
translating the evaluation samples into a source language (usually English) and
using monolingual models (i.e., translate-test). However, our analysis reveals
that translated texts contain unique characteristics distinct from
human-written ones, referred to as translation artifacts. We find that these
artifacts can significantly affect the models, confirmed by extensive
experiments across diverse models, languages, and translation processes. In
light of this, we present a simple data augmentation strategy that can
alleviate the adverse impacts of translation artifacts. | cs.CL | Translation Deserves Better: Analyzing Translation Artifacts in Cross-lingual Visual Question Answering |
2024-06-04T13:58:28Z | http://arxiv.org/pdf/2406.02329v1 | http://arxiv.org/abs/2406.02329v1 | Robin SM Chan, Reda Boumasmoud, Anej Svete, Yuxin Ren, Qipeng Guo, Zhijing Jin, Shauli Ravfogel, Mrinmaya Sachan, Bernhard Schölkopf, Mennatallah El-Assady, Ryan Cotterell | 2024-06-04T13:58:28Z | cs.CL, cs.LG | Pre-trained language encoders -- functions that represent text as vectors --
are an integral component of many NLP tasks. We tackle a natural question in
language encoder analysis: What does it mean for two encoders to be similar? We
contend that a faithful measure of similarity needs to be \emph{intrinsic},
that is, task-independent, yet still be informative of \emph{extrinsic}
similarity -- the performance on downstream tasks. It is common to consider two
encoders similar if they are \emph{homotopic}, i.e., if they can be aligned
through some transformation. In this spirit, we study the properties of
\emph{affine} alignment of language encoders and its implications on extrinsic
similarity. We find that while affine alignment is fundamentally an asymmetric
notion of similarity, it is still informative of extrinsic similarity. We
confirm this on datasets of natural language representations. Beyond providing
useful bounds on extrinsic similarity, affine intrinsic similarity also allows
us to begin uncovering the structure of the space of pre-trained encoders by
defining an order over them. | cs.CL | On Affine Homotopy between Language Encoders |
2024-06-04T13:57:22Z | http://arxiv.org/pdf/2406.02325v1 | http://arxiv.org/abs/2406.02325v1 | Felipe A. Rodriguez Y. | 2024-06-04T13:57:22Z | cs.CL, cs.AI | Large Language Models (LLMs) are continuously being applied in a more diverse
set of contexts. At their current state, however, even state-of-the-art LLMs
such as Generative Pre-Trained Transformer 4 (GTP-4) have challenges when
extracting information from real-world technical documentation without a heavy
preprocessing. One such area with real-world technical documentation is
telecommunications engineering, which could greatly benefit from
domain-specific LLMs. The unique format and overall structure of
telecommunications internal specifications differs greatly from standard
English and thus it is evident that the application of out-of-the-box Natural
Language Processing (NLP) tools is not a viable option. In this article, we
outline the limitations of out-of-the-box NLP tools for processing technical
information generated by telecommunications experts, and expand the concept of
Technical Language Processing (TLP) to the telecommunication domain.
Additionally, we explore the effect of domain-specific LLMs in the work of
Specification Engineers, emphasizing the potential benefits of adopting
domain-specific LLMs to speed up the training of experts in different
telecommunications fields. | cs.CL | Technical Language Processing for Telecommunications Specifications |
2024-06-04T13:52:54Z | http://arxiv.org/pdf/2406.06579v2 | http://arxiv.org/abs/2406.06579v2 | Xiaofeng Zhang, Chen Shen, Xiaosong Yuan, Shaotian Yan, Liang Xie, Wenxiao Wang, Chaochen Gu, Hao Tang, Jieping Ye | 2024-06-13T10:29:45Z | cs.CL, cs.AI, cs.CV | Recently, multimodal large language models have exploded with an endless
variety, most of the popular Large Vision Language Models (LVLMs) depend on
sequential visual representation, where images are converted into hundreds or
thousands of tokens before being input into the Large Language Model (LLM)
along with language prompts. The black-box design hinders the interpretability
of visual-language models, especially regarding more complex reasoning tasks.
To explore the interaction process between image and text in complex reasoning
tasks, we introduce the information flow method to visualize the interaction
mechanism. By analyzing the dynamic flow of the information flow, we find that
the information flow appears to converge in the shallow layer. Further
investigation revealed a redundancy of the image token in the shallow layer.
Consequently, a truncation strategy was introduced to aggregate image tokens
within these shallow layers. This approach has been validated through
experiments across multiple models, yielding consistent improvements. | cs.CL | From Redundancy to Relevance: Enhancing Explainability in Multimodal Large Language Models |
2024-06-04T13:44:36Z | http://arxiv.org/pdf/2406.06578v1 | http://arxiv.org/abs/2406.06578v1 | Dare Azeez Oyeyemi, Adebola K. Ojo | 2024-06-04T13:44:36Z | cs.CL, cs.AI, cs.LG | In the modern era, mobile phones have become ubiquitous, and Short Message
Service (SMS) has grown to become a multi-million-dollar service due to the
widespread adoption of mobile devices and the millions of people who use SMS
daily. However, SMS spam has also become a pervasive problem that endangers
users' privacy and security through phishing and fraud. Despite numerous spam
filtering techniques, there is still a need for a more effective solution to
address this problem [1]. This research addresses the pervasive issue of SMS
spam, which poses threats to users' privacy and security. Despite existing spam
filtering techniques, the high false-positive rate persists as a challenge. The
study introduces a novel approach utilizing Natural Language Processing (NLP)
and machine learning models, particularly BERT (Bidirectional Encoder
Representations from Transformers), for SMS spam detection and classification.
Data preprocessing techniques, such as stop word removal and tokenization, are
applied, along with feature extraction using BERT. Machine learning models,
including SVM, Logistic Regression, Naive Bayes, Gradient Boosting, and Random
Forest, are integrated with BERT for differentiating spam from ham messages.
Evaluation results revealed that the Na\"ive Bayes classifier + BERT model
achieves the highest accuracy at 97.31% with the fastest execution time of 0.3
seconds on the test dataset. This approach demonstrates a notable enhancement
in spam detection efficiency and a low false-positive rate. The developed model
presents a valuable solution to combat SMS spam, ensuring faster and more
accurate detection. This model not only safeguards users' privacy but also
assists network providers in effectively identifying and blocking SMS spam
messages. | cs.CL | SMS Spam Detection and Classification to Combat Abuse in Telephone Networks Using Natural Language Processing |
2024-06-04T13:30:45Z | http://arxiv.org/pdf/2406.02301v1 | http://arxiv.org/abs/2406.02301v1 | Huiyuan Lai, Malvina Nissim | 2024-06-04T13:30:45Z | cs.CL | Large language models (LLMs) with Chain-of-thought (CoT) have recently
emerged as a powerful technique for eliciting reasoning to improve various
downstream tasks. As most research mainly focuses on English, with few
explorations in a multilingual context, the question of how reliable this
reasoning capability is in different languages is still open. To address it
directly, we study multilingual reasoning consistency across multiple
languages, using popular open-source LLMs. First, we compile the first
large-scale multilingual math reasoning dataset, mCoT-MATH, covering eleven
diverse languages. Then, we introduce multilingual CoT instruction tuning to
boost reasoning capability across languages, thereby improving model
consistency. While existing LLMs show substantial variation across the
languages we consider, and especially low performance for lesser resourced
languages, our 7B parameter model mCoT achieves impressive consistency across
languages, and superior or comparable performance to close- and open-source
models even of much larger sizes. | cs.CL | mCoT: Multilingual Instruction Tuning for Reasoning Consistency in Language Models |
2024-06-04T12:43:47Z | http://arxiv.org/pdf/2406.02267v1 | http://arxiv.org/abs/2406.02267v1 | Nathaniel Berger, Stefan Riezler, Miriam Exel, Matthias Huck | 2024-06-04T12:43:47Z | cs.CL | While large language models (LLMs) pre-trained on massive amounts of unpaired
language data have reached the state-of-the-art in machine translation (MT) of
general domain texts, post-editing (PE) is still required to correct errors and
to enhance term translation quality in specialized domains. In this paper we
present a pilot study of enhancing translation memories (TM) produced by PE
(source segments, machine translations, and reference translations, henceforth
called PE-TM) for the needs of correct and consistent term translation in
technical domains.
We investigate a light-weight two-step scenario where, at inference time, a
human translator marks errors in the first translation step, and in a second
step a few similar examples are extracted from the PE-TM to prompt an LLM. Our
experiment shows that the additional effort of augmenting translations with
human error markings guides the LLM to focus on a correction of the marked
errors, yielding consistent improvements over automatic PE (APE) and MT from
scratch. | cs.CL | Prompting Large Language Models with Human Error Markings for Self-Correcting Machine Translation |
2024-06-04T12:43:23Z | http://arxiv.org/pdf/2406.02266v1 | http://arxiv.org/abs/2406.02266v1 | Chuankai Xu, Dongming Zhao, Bo Wang, Hanwen Xing | 2024-06-04T12:43:23Z | cs.CL | Despite the prevalence of retrieval-augmented language models (RALMs), the
seamless integration of these models with retrieval mechanisms to enhance
performance in document-based tasks remains challenging. While some
post-retrieval processing Retrieval-Augmented Generation (RAG) methods have
achieved success, most still lack the ability to distinguish pertinent from
extraneous information, leading to potential inconsistencies and reduced
precision in the generated output, which subsequently affects the truthfulness
of the language model's responses. To address these limitations, this work
proposes a novel two-stage consistency learning approach for retrieved
information compression in retrieval-augmented language models to enhance
performance. By incorporating consistency learning, the aim is to generate
summaries that maintain coherence and alignment with the intended semantic
representations of a teacher model while improving faithfulness to the original
retrieved documents. The proposed method is empirically validated across
multiple datasets, demonstrating notable enhancements in precision and
efficiency for question-answering tasks. It outperforms existing baselines and
showcases the synergistic effects of combining contrastive and consistency
learning paradigms within the retrieval-augmented generation framework. | cs.CL | Enhancing Retrieval-Augmented LMs with a Two-stage Consistency Learning Compressor |
2024-06-04T12:41:54Z | http://arxiv.org/pdf/2406.02265v2 | http://arxiv.org/abs/2406.02265v2 | Wenyan Li, Jiaang Li, Rita Ramos, Raphael Tang, Desmond Elliott | 2024-06-06T13:59:03Z | cs.CV, cs.CL | Recent advances in retrieval-augmented models for image captioning highlight
the benefit of retrieving related captions for efficient, lightweight models
with strong domain-transfer capabilities. While these models demonstrate the
success of retrieval augmentation, retrieval models are still far from perfect
in practice: the retrieved information can sometimes mislead the model,
resulting in incorrect generation and worse performance. In this paper, we
analyze the robustness of a retrieval-augmented captioning model SmallCap. Our
analysis shows that the model is sensitive to tokens that appear in the
majority of the retrieved captions, and the input attribution shows that those
tokens are likely copied into the generated output. Given these findings, we
propose to train the model by sampling retrieved captions from more diverse
sets. This decreases the chance that the model learns to copy majority tokens,
and improves both in-domain and cross-domain performance. | cs.CV | Understanding Retrieval Robustness for Retrieval-Augmented Image Captioning |
2024-06-04T12:17:16Z | http://arxiv.org/pdf/2406.02251v1 | http://arxiv.org/abs/2406.02251v1 | Lukas Christ, Shahin Amiriparian, Manuel Milling, Ilhan Aslan, Björn W. Schuller | 2024-06-04T12:17:16Z | cs.CL, cs.AI | Telling stories is an integral part of human communication which can evoke
emotions and influence the affective states of the audience. Automatically
modeling emotional trajectories in stories has thus attracted considerable
scholarly interest. However, as most existing works have been limited to
unsupervised dictionary-based approaches, there is no benchmark for this task.
We address this gap by introducing continuous valence and arousal labels for an
existing dataset of children's stories originally annotated with discrete
emotion categories. We collect additional annotations for this data and map the
categorical labels to the continuous valence and arousal space. For predicting
the thus obtained emotionality signals, we fine-tune a DeBERTa model and
improve upon this baseline via a weakly supervised learning approach. The best
configuration achieves a Concordance Correlation Coefficient (CCC) of $.8221$
for valence and $.7125$ for arousal on the test set, demonstrating the efficacy
of our proposed approach. A detailed analysis shows the extent to which the
results vary depending on factors such as the author, the individual story, or
the section within the story. In addition, we uncover the weaknesses of our
approach by investigating examples that prove to be difficult to predict. | cs.CL | Modeling Emotional Trajectories in Written Stories Utilizing Transformers and Weakly-Supervised Learning |
2024-06-04T12:09:44Z | http://arxiv.org/pdf/2406.02245v1 | http://arxiv.org/abs/2406.02245v1 | Gabriele Picco, Leopold Fuchs, Marcos Martínez Galindo, Alberto Purpura, Vanessa López, Hoang Thanh Lam | 2024-06-04T12:09:44Z | cs.CL, cs.IR, cs.LG | Zero-shot entity and relation classification models leverage available
external information of unseen classes -- e.g., textual descriptions -- to
annotate input text data. Thanks to the minimum data requirement, Zero-Shot
Learning (ZSL) methods have high value in practice, especially in applications
where labeled data is scarce. Even though recent research in ZSL has
demonstrated significant results, our analysis reveals that those methods are
sensitive to provided textual descriptions of entities (or relations). Even a
minor modification of descriptions can lead to a change in the decision
boundary between entity (or relation) classes. In this paper, we formally
define the problem of identifying effective descriptions for zero shot
inference. We propose a strategy for generating variations of an initial
description, a heuristic for ranking them and an ensemble method capable of
boosting the predictions of zero-shot models through description enhancement.
Empirical results on four different entity and relation classification datasets
show that our proposed method outperform existing approaches and achieve new
SOTA results on these datasets under the ZSL settings. The source code of the
proposed solutions and the evaluation framework are open-sourced. | cs.CL | Description Boosting for Zero-Shot Entity and Relation Classification |
2024-06-04T11:57:58Z | http://arxiv.org/pdf/2406.02237v1 | http://arxiv.org/abs/2406.02237v1 | Donglei Yu, Xiaomian Kang, Yuchen Liu, Yu Zhou, Chengqing Zong | 2024-06-04T11:57:58Z | cs.CL | Simultaneous Machine Translation (SiMT) generates target outputs while
receiving stream source inputs and requires a read/write policy to decide
whether to wait for the next source token or generate a new target token, whose
decisions form a \textit{decision path}. Existing SiMT methods, which learn the
policy by exploring various decision paths in training, face inherent
limitations. These methods not only fail to precisely optimize the policy due
to the inability to accurately assess the individual impact of each decision on
SiMT performance, but also cannot sufficiently explore all potential paths
because of their vast number. Besides, building decision paths requires
unidirectional encoders to simulate streaming source inputs, which impairs the
translation quality of SiMT models. To solve these issues, we propose
\textbf{S}elf-\textbf{M}odifying \textbf{S}tate \textbf{M}odeling (SM$^2$), a
novel training paradigm for SiMT task. Without building decision paths, SM$^2$
individually optimizes decisions at each state during training. To precisely
optimize the policy, SM$^2$ introduces Self-Modifying process to independently
assess and adjust decisions at each state. For sufficient exploration, SM$^2$
proposes Prefix Sampling to efficiently traverse all potential states.
Moreover, SM$^2$ ensures compatibility with bidirectional encoders, thus
achieving higher translation quality. Experiments show that SM$^2$ outperforms
strong baselines. Furthermore, SM$^2$ allows offline machine translation models
to acquire SiMT ability with fine-tuning. | cs.CL | Self-Modifying State Modeling for Simultaneous Machine Translation |
2024-06-04T11:36:09Z | http://arxiv.org/pdf/2406.02224v2 | http://arxiv.org/abs/2406.02224v2 | Tao Fan, Guoqiang Ma, Yan Kang, Hanlin Gu, Yuanfeng Song, Lixin Fan, Kai Chen, Qiang Yang | 2024-06-18T08:17:00Z | cs.CL, cs.AI | Recent research in federated large language models (LLMs) has primarily
focused on enabling clients to fine-tune their locally deployed homogeneous
LLMs collaboratively or on transferring knowledge from server-based LLMs to
small language models (SLMs) at downstream clients. However, a significant gap
remains in the simultaneous mutual enhancement of both the server's LLM and
clients' SLMs. To bridge this gap, we propose FedMKT, a parameter-efficient
federated mutual knowledge transfer framework for large and small language
models. This framework is designed to adaptively transfer knowledge from the
server's LLM to clients' SLMs while concurrently enriching the LLM with
clients' unique domain insights. We facilitate token alignment using minimum
edit distance (MinED) and then selective mutual knowledge transfer between
client-side SLMs and a server-side LLM, aiming to collectively enhance their
performance. Through extensive experiments across three distinct scenarios, we
evaluate the effectiveness of FedMKT using various public LLMs and SLMs on a
range of NLP text generation tasks. Empirical results demonstrate that FedMKT
simultaneously boosts the performance of both LLMs and SLMs. | cs.CL | FedMKT: Federated Mutual Knowledge Transfer for Large and Small Language Models |
2024-06-04T11:06:13Z | http://arxiv.org/pdf/2406.02208v1 | http://arxiv.org/abs/2406.02208v1 | Haodong Hong, Sen Wang, Zi Huang, Qi Wu, Jiajun Liu | 2024-06-04T11:06:13Z | cs.CV, cs.AI, cs.CL | Current Vision-and-Language Navigation (VLN) tasks mainly employ textual
instructions to guide agents. However, being inherently abstract, the same
textual instruction can be associated with different visual signals, causing
severe ambiguity and limiting the transfer of prior knowledge in the vision
domain from the user to the agent. To fill this gap, we propose
Vision-and-Language Navigation with Multi-modal Prompts (VLN-MP), a novel task
augmenting traditional VLN by integrating both natural language and images in
instructions. VLN-MP not only maintains backward compatibility by effectively
handling text-only prompts but also consistently shows advantages with
different quantities and relevance of visual prompts. Possible forms of visual
prompts include both exact and similar object images, providing adaptability
and versatility in diverse navigation scenarios. To evaluate VLN-MP under a
unified framework, we implement a new benchmark that offers: (1) a
training-free pipeline to transform textual instructions into multi-modal forms
with landmark images; (2) diverse datasets with multi-modal instructions for
different downstream tasks; (3) a novel module designed to process various
image prompts for seamless integration with state-of-the-art VLN models.
Extensive experiments on four VLN benchmarks (R2R, RxR, REVERIE, CVDN) show
that incorporating visual prompts significantly boosts navigation performance.
While maintaining efficiency with text-only prompts, VLN-MP enables agents to
navigate in the pre-explore setting and outperform text-based models, showing
its broader applicability. | cs.CV | Why Only Text: Empowering Vision-and-Language Navigation with Multi-modal Prompts |
2024-06-04T09:58:29Z | http://arxiv.org/pdf/2406.02169v2 | http://arxiv.org/abs/2406.02169v2 | Saminu Mohammad Aliyu, Gregory Maksha Wajiga, Muhammad Murtala | 2024-06-05T20:07:50Z | cs.CL, 14F05, F.2.2; I.2.7 | The proliferation of online offensive language necessitates the development
of effective detection mechanisms, especially in multilingual contexts. This
study addresses the challenge by developing and introducing novel datasets for
offensive language detection in three major Nigerian languages: Hausa, Yoruba,
and Igbo. We collected data from Twitter and manually annotated it to create
datasets for each of the three languages, using native speakers. We used
pre-trained language models to evaluate their efficacy in detecting offensive
language in our datasets. The best-performing model achieved an accuracy of
90\%. To further support research in offensive language detection, we plan to
make the dataset and our models publicly available. | cs.CL | A multilingual dataset for offensive language and hate speech detection for hausa, yoruba and igbo languages |
2024-06-04T09:56:05Z | http://arxiv.org/pdf/2406.02166v1 | http://arxiv.org/abs/2406.02166v1 | Saierdaer Yusuyin, Te Ma, Hao Huang, Wenbo Zhao, Zhijian Ou | 2024-06-04T09:56:05Z | cs.SD, cs.CL, eess.AS | There exist three approaches for multilingual and crosslingual automatic
speech recognition (MCL-ASR) - supervised pre-training with phonetic or
graphemic transcription, and self-supervised pre-training. We find that
pre-training with phonetic supervision has been underappreciated so far for
MCL-ASR, while conceptually it is more advantageous for information sharing
between different languages. This paper explores the approach of pre-training
with weakly phonetic supervision towards data-efficient MCL-ASR, which is
called Whistle. We relax the requirement of gold-standard human-validated
phonetic transcripts, and obtain International Phonetic Alphabet (IPA) based
transcription by leveraging the LanguageNet grapheme-to-phoneme (G2P) models.
We construct a common experimental setup based on the CommonVoice dataset,
called CV-Lang10, with 10 seen languages and 2 unseen languages. A set of
experiments are conducted on CV-Lang10 to compare, as fair as possible, the
three approaches under the common setup for MCL-ASR. Experiments demonstrate
the advantages of phoneme-based models (Whistle) for MCL-ASR, in terms of
speech recognition for seen languages, crosslingual performance for unseen
languages with different amounts of few-shot data, overcoming catastrophic
forgetting, and training efficiency.It is found that when training data is more
limited, phoneme supervision can achieve better results compared to subword
supervision and self-supervision, thereby providing higher data-efficiency. To
support reproducibility and promote future research along this direction, we
will release the code, models and data for the whole pipeline of Whistle at
https://github.com/thu-spmi/CAT upon publication. | cs.SD | Whistle: Data-Efficient Multilingual and Crosslingual Speech Recognition via Weakly Phonetic Supervision |
2024-06-04T09:35:47Z | http://arxiv.org/pdf/2406.02148v1 | http://arxiv.org/abs/2406.02148v1 | Qingkai Min, Qipeng Guo, Xiangkun Hu, Songfang Huang, Zheng Zhang, Yue Zhang | 2024-06-04T09:35:47Z | cs.CL, cs.AI | Cross-document event coreference resolution (CDECR) involves clustering event
mentions across multiple documents that refer to the same real-world events.
Existing approaches utilize fine-tuning of small language models (SLMs) like
BERT to address the compatibility among the contexts of event mentions.
However, due to the complexity and diversity of contexts, these models are
prone to learning simple co-occurrences. Recently, large language models (LLMs)
like ChatGPT have demonstrated impressive contextual understanding, yet they
encounter challenges in adapting to specific information extraction (IE) tasks.
In this paper, we propose a collaborative approach for CDECR, leveraging the
capabilities of both a universally capable LLM and a task-specific SLM. The
collaborative strategy begins with the LLM accurately and comprehensively
summarizing events through prompting. Then, the SLM refines its learning of
event representations based on these insights during fine-tuning. Experimental
results demonstrate that our approach surpasses the performance of both the
large and small language models individually, forming a complementary
advantage. Across various datasets, our approach achieves state-of-the-art
performance, underscoring its effectiveness in diverse scenarios. | cs.CL | Synergetic Event Understanding: A Collaborative Approach to Cross-Document Event Coreference Resolution with Large Language Models |
2024-06-04T09:31:18Z | http://arxiv.org/pdf/2406.02143v1 | http://arxiv.org/abs/2406.02143v1 | Ruichao Yang, Wei Gao, Jing Ma, Hongzhan Lin, Bo Wang | 2024-06-04T09:31:18Z | cs.CL | Learning multi-task models for jointly detecting stance and verifying rumors
poses challenges due to the need for training data of stance at post level and
rumor veracity at claim level, which are difficult to obtain. To address this
issue, we leverage large language models (LLMs) as the foundation annotators
for the joint stance detection (SD) and rumor verification (RV) tasks, dubbed
as JSDRV. We introduce a novel reinforcement tuning framework to enhance the
joint predictive capabilities of LLM-based SD and RV components. Specifically,
we devise a policy for selecting LLM-annotated data at the two levels,
employing a hybrid reward mechanism to choose high-quality labels for effective
LLM fine-tuning on both tasks. Results demonstrate that JSDRV improves the
capabilities of LLMs in the joint tasks, not only outperforming
state-of-the-art methods but also generalizing to non-LLMs accommodated as task
models. | cs.CL | Reinforcement Tuning for Detecting Stances and Debunking Rumors Jointly with Large Language Models |
2024-06-04T09:24:04Z | http://arxiv.org/pdf/2406.02135v1 | http://arxiv.org/abs/2406.02135v1 | Ben Chen, Huangyu Dai, Xiang Ma, Wen Jiang, Wei Ning | 2024-06-04T09:24:04Z | cs.IR, cs.CL | Semantic relevance calculation is crucial for e-commerce search engines, as
it ensures that the items selected closely align with customer intent.
Inadequate attention to this aspect can detrimentally affect user experience
and engagement. Traditional text-matching techniques are prevalent but often
fail to capture the nuances of search intent accurately, so neural networks now
have become a preferred solution to processing such complex text matching.
Existing methods predominantly employ representation-based architectures, which
strike a balance between high traffic capacity and low latency. However, they
exhibit significant shortcomings in generalization and robustness when compared
to interaction-based architectures. In this work, we introduce a robust
interaction-based modeling paradigm to address these shortcomings. It
encompasses 1) a dynamic length representation scheme for expedited inference,
2) a professional terms recognition method to identify subjects and core
attributes from complex sentence structures, and 3) a contrastive adversarial
training protocol to bolster the model's robustness and matching capabilities.
Extensive offline evaluations demonstrate the superior robustness and
effectiveness of our approach, and online A/B testing confirms its ability to
improve relevance in the same exposure position, resulting in more clicks and
conversions. To the best of our knowledge, this method is the first
interaction-based approach for large e-commerce search relevance calculation.
Notably, we have deployed it for the entire search traffic on alibaba.com, the
largest B2B e-commerce platform in the world. | cs.IR | Robust Interaction-based Relevance Modeling for Online E-Commerce and LLM-based Retrieval |
2024-06-04T09:23:30Z | http://arxiv.org/pdf/2406.02134v1 | http://arxiv.org/abs/2406.02134v1 | Danqing Hu, Shanyuan Zhang, Qing Liu, Xiaofeng Zhu, Bing Liu | 2024-06-04T09:23:30Z | cs.CL | Large language models (LLMs) like ChatGPT show excellent capabilities in
various natural language processing tasks, especially for text generation. The
effectiveness of LLMs in summarizing radiology report impressions remains
unclear. In this study, we explore the capability of eight LLMs on the
radiology report impression summarization. Three types of radiology reports,
i.e., CT, PET-CT, and Ultrasound reports, are collected from Peking University
Cancer Hospital and Institute. We use the report findings to construct the
zero-shot, one-shot, and three-shot prompts with complete example reports to
generate the impressions. Besides the automatic quantitative evaluation
metrics, we define five human evaluation metrics, i.e., completeness,
correctness, conciseness, verisimilitude, and replaceability, to evaluate the
semantics of the generated impressions. Two thoracic surgeons (ZSY and LB) and
one radiologist (LQ) compare the generated impressions with the reference
impressions and score each impression under the five human evaluation metrics.
Experimental results show that there is a gap between the generated impressions
and reference impressions. Although the LLMs achieve comparable performance in
completeness and correctness, the conciseness and verisimilitude scores are not
very high. Using few-shot prompts can improve the LLMs' performance in
conciseness and verisimilitude, but the clinicians still think the LLMs can not
replace the radiologists in summarizing the radiology impressions. | cs.CL | The current status of large language models in summarizing radiology report impressions |
2024-06-04T09:21:31Z | http://arxiv.org/pdf/2406.02133v1 | http://arxiv.org/abs/2406.02133v1 | Alex Agranovich, Eliya Nachmani, Oleg Rybakov, Yifan Ding, Ye Jia, Nadav Bar, Heiga Zen, Michelle Tadmor Ramanovich | 2024-06-04T09:21:31Z | eess.AS, cs.CL, cs.LG, cs.SD | Simultaneous speech-to-speech translation (S2ST) holds the promise of
breaking down communication barriers and enabling fluid conversations across
languages. However, achieving accurate, real-time translation through mobile
devices remains a major challenge. We introduce SimulTron, a novel S2ST
architecture designed to tackle this task. SimulTron is a lightweight direct
S2ST model that uses the strengths of the Translatotron framework while
incorporating key modifications for streaming operation, and an adjustable
fixed delay. Our experiments show that SimulTron surpasses Translatotron 2 in
offline evaluations. Furthermore, real-time evaluations reveal that SimulTron
improves upon the performance achieved by Translatotron 1. Additionally,
SimulTron achieves superior BLEU scores and latency compared to previous
real-time S2ST method on the MuST-C dataset. Significantly, we have
successfully deployed SimulTron on a Pixel 7 Pro device, show its potential for
simultaneous S2ST on-device. | eess.AS | SimulTron: On-Device Simultaneous Speech to Speech Translation |
2024-06-04T09:11:46Z | http://arxiv.org/pdf/2406.02128v1 | http://arxiv.org/abs/2406.02128v1 | Vivien Cabannes, Charles Arnal, Wassim Bouaziz, Alice Yang, Francois Charton, Julia Kempe | 2024-06-04T09:11:46Z | cs.LG, cs.AI, cs.CL | Chain-of-Thought (CoT) reasoning is known to improve Large Language Models
both empirically and in terms of theoretical approximation power. However, our
understanding of the inner workings and conditions of apparition of CoT
capabilities remains limited. This paper helps fill this gap by demonstrating
how CoT reasoning emerges in transformers in a controlled and interpretable
setting. In particular, we observe the appearance of a specialized attention
mechanism dedicated to iterative reasoning, which we coined "iteration heads".
We track both the emergence and the precise working of these iteration heads
down to the attention level, and measure the transferability of the CoT skills
to which they give rise between tasks. | cs.LG | Iteration Head: A Mechanistic Study of Chain-of-Thought |
2024-06-04T09:02:22Z | http://arxiv.org/pdf/2406.02120v1 | http://arxiv.org/abs/2406.02120v1 | Jinliang Lu, Chen Wang, Jiajun Zhang | 2024-06-04T09:02:22Z | cs.CL | Large language models (LLMs) have shown impressive capabilities in adapting
to various tasks when provided with task-specific instructions. However, LLMs
using standard decoding strategies often struggle with deviations from the
inputs. Intuitively, compliant LLM outputs should reflect the information
present in the input, which can be measured by point-wise mutual information
(PMI) scores. Therefore, we propose Diver, a novel approach that enhances LLM
Decoding through span-level PMI verification. During inference, Diver first
identifies divergence steps that may lead to multiple candidate spans.
Subsequently, it calculates the PMI scores by assessing the log-likelihood
gains of the input if the candidate spans are generated. Finally, the optimal
span is selected based on the PMI re-ranked output distributions. We evaluate
our method across various downstream tasks, and empirical results demonstrate
that Diver significantly outperforms existing decoding methods in both
performance and versatility. | cs.CL | Diver: Large Language Model Decoding with Span-Level Mutual Information Verification |
2024-06-04T08:36:39Z | http://arxiv.org/pdf/2406.02110v1 | http://arxiv.org/abs/2406.02110v1 | Zhuoyang Li, Liran Deng, Hui Liu, Qiaoqiao Liu, Junzhao Du | 2024-06-04T08:36:39Z | cs.CL, cs.AI | OwnThink stands as the most extensive Chinese open-domain knowledge graph
introduced in recent times. Despite prior attempts in question answering over
OwnThink (OQA), existing studies have faced limitations in model representation
capabilities, posing challenges in further enhancing overall accuracy in
question answering. In this paper, we introduce UniOQA, a unified framework
that integrates two complementary parallel workflows. Unlike conventional
approaches, UniOQA harnesses large language models (LLMs) for precise question
answering and incorporates a direct-answer-prediction process as a
cost-effective complement. Initially, to bolster representation capacity, we
fine-tune an LLM to translate questions into the Cypher query language (CQL),
tackling issues associated with restricted semantic understanding and
hallucinations. Subsequently, we introduce the Entity and Relation Replacement
algorithm to ensure the executability of the generated CQL. Concurrently, to
augment overall accuracy in question answering, we further adapt the
Retrieval-Augmented Generation (RAG) process to the knowledge graph.
Ultimately, we optimize answer accuracy through a dynamic decision algorithm.
Experimental findings illustrate that UniOQA notably advances SpCQL Logical
Accuracy to 21.2% and Execution Accuracy to 54.9%, achieving the new
state-of-the-art results on this benchmark. Through ablation experiments, we
delve into the superior representation capacity of UniOQA and quantify its
performance breakthrough. | cs.CL | UniOQA: A Unified Framework for Knowledge Graph Question Answering with Large Language Models |
2024-06-04T08:35:04Z | http://arxiv.org/pdf/2406.02106v1 | http://arxiv.org/abs/2406.02106v1 | Weiqi Wang, Yangqiu Song | 2024-06-04T08:35:04Z | cs.CL | To enable Large Language Models (LLMs) to function as conscious agents with
generalizable reasoning capabilities, it is crucial that they possess the
reasoning ability to comprehend situational changes (transitions) in
distribution triggered by environmental factors or actions from other agents.
Despite its fundamental significance, this ability remains underexplored due to
the complexity of modeling infinite possible changes in an event and their
associated distributions, coupled with the lack of benchmark data with
situational transitions. Addressing these gaps, we propose a novel formulation
of reasoning with distributional changes as a three-step discriminative
process, termed as MetAphysical ReaSoning. We then introduce the first-ever
benchmark, MARS, comprising three tasks corresponding to each step. These tasks
systematically assess LLMs' capabilities in reasoning the plausibility of (i)
changes in actions, (ii) states caused by changed actions, and (iii)
situational transitions driven by changes in action. Extensive evaluations with
20 (L)LMs of varying sizes and methods indicate that all three tasks in this
process pose significant challenges, even for state-of-the-art LLMs and LMs
after fine-tuning. Further analyses reveal potential causes for the
underperformance of LLMs and demonstrate that pre-training them on large-scale
conceptualization taxonomies can potentially enhance their metaphysical
reasoning capabilities. Our data and models are publicly accessible at
https://github.com/HKUST-KnowComp/MARS. | cs.CL | MARS: Benchmarking the Metaphysical Reasoning Abilities of Language Models with a Multi-task Evaluation Dataset |
2024-06-04T08:34:19Z | http://arxiv.org/pdf/2406.06577v1 | http://arxiv.org/abs/2406.06577v1 | Jing Yang, Xiao Wang, Yu Zhao, Yuhang Liu, Fei-Yue Wang | 2024-06-04T08:34:19Z | cs.CL, cs.AI | Crowdsourcing is a critical technology in social manufacturing, which
leverages an extensive and boundless reservoir of human resources to handle a
wide array of complex tasks. The successful execution of these complex tasks
relies on task decomposition (TD) and allocation, with the former being a
prerequisite for the latter. Recently, pre-trained language models (PLMs)-based
methods have garnered significant attention. However, they are constrained to
handling straightforward common-sense tasks due to their inherent restrictions
involving limited and difficult-to-update knowledge as well as the presence of
hallucinations. To address these issues, we propose a retrieval-augmented
generation-based crowdsourcing framework that reimagines TD as event detection
from the perspective of natural language understanding. However, the existing
detection methods fail to distinguish differences between event types and
always depend on heuristic rules and external semantic analyzing tools.
Therefore, we present a Prompt-Based Contrastive learning framework for TD
(PBCT), which incorporates a prompt-based trigger detector to overcome
dependence. Additionally, trigger-attentive sentinel and masked contrastive
learning are introduced to provide varying attention to trigger and contextual
features according to different event types. Experiment results demonstrate the
competitiveness of our method in both supervised and zero-shot detection. A
case study on printed circuit board manufacturing is showcased to validate its
adaptability to unknown professional domains. | cs.CL | RAG-based Crowdsourcing Task Decomposition via Masked Contrastive Learning with Prompts |
2024-06-04T08:30:37Z | http://arxiv.org/pdf/2406.02100v1 | http://arxiv.org/abs/2406.02100v1 | Haolong Li, Yu Ma, Yinqi Zhang, Chen Ye, Jie Chen | 2024-06-04T08:30:37Z | cs.CL | Large Language Models (LLMs) have shown excellent performance in language
understanding, text generation, code synthesis, and many other tasks, while
they still struggle in complex multi-step reasoning problems, such as
mathematical reasoning. In this paper, through a newly proposed arithmetical
puzzle problem, we show that the model can perform well on multi-step reasoning
tasks via fine-tuning on high-quality synthetic data. Experimental results with
the open-llama-3B model on three different test datasets show that not only the
model can reach a zero-shot pass@1 at 0.44 on the in-domain dataset, it also
demonstrates certain generalization capabilities on the out-of-domain datasets.
Specifically, this paper has designed two out-of-domain datasets in the form of
extending the numerical range and the composing components of the arithmetical
puzzle problem separately. The fine-tuned models have shown encouraging
performance on these two far more difficult tasks with the zero-shot pass@1 at
0.33 and 0.35, respectively. | cs.CL | Exploring Mathematical Extrapolation of Large Language Models with Synthetic Data |
2024-06-04T08:02:39Z | http://arxiv.org/pdf/2406.02080v1 | http://arxiv.org/abs/2406.02080v1 | Shida Wang | 2024-06-04T08:02:39Z | cs.CL, cs.AI, cs.LG, math.DS | In this paper, we investigate the length-extension of state-space models
(SSMs) in language modeling. Length extension involves training models on short
sequences and testing them on longer ones. We show that state-space models
trained with zero hidden states initialization have difficulty doing length
extension. We explain this difficulty by pointing out the length extension is
equivalent to polynomial extrapolation. Based on the theory, we propose a
simple yet effective method - changing the hidden states initialization scheme
- to improve the length extension. Moreover, our method shows that using long
training sequence length is beneficial but not necessary to length extension.
Changing the hidden state initialization enables the efficient training of
long-memory model with a smaller training context length. | cs.CL | LongSSM: On the Length Extension of State-space Models in Language Modelling |
2024-06-04T08:00:40Z | http://arxiv.org/pdf/2406.02079v1 | http://arxiv.org/abs/2406.02079v1 | Yida Cai, Hao Sun, Hsiu-Yuan Huang, Yunfang Wu | 2024-06-04T08:00:40Z | cs.CL | Information Extraction (IE) plays a crucial role in Natural Language
Processing (NLP) by extracting structured information from unstructured text,
thereby facilitating seamless integration with various real-world applications
that rely on structured data. Despite its significance, recent experiments
focusing on English IE tasks have shed light on the challenges faced by Large
Language Models (LLMs) in achieving optimal performance, particularly in
sub-tasks like Named Entity Recognition (NER). In this paper, we delve into a
comprehensive investigation of the performance of mainstream Chinese
open-source LLMs in tackling IE tasks, specifically under zero-shot conditions
where the models are not fine-tuned for specific tasks. Additionally, we
present the outcomes of several few-shot experiments to further gauge the
capability of these models. Moreover, our study includes a comparative analysis
between these open-source LLMs and ChatGPT, a widely recognized language model,
on IE performance. Through meticulous experimentation and analysis, we aim to
provide insights into the strengths, limitations, and potential enhancements of
existing Chinese open-source LLMs in the domain of Information Extraction
within the context of NLP. | cs.CL | Assessing the Performance of Chinese Open Source Large Language Models in Information Extraction Tasks |
2024-06-04T07:51:30Z | http://arxiv.org/pdf/2406.02069v2 | http://arxiv.org/abs/2406.02069v2 | Zefan Cai., Yichi Zhang, Bofei Gao, Yuliang Liu, Tianyu Liu, Keming Lu, Wayne Xiong, Yue Dong, Baobao Chang, Junjie Hu, Wen Xiao | 2024-06-16T06:41:08Z | cs.CL, cs.AI | In this study, we investigate whether attention-based information flow inside
large language models (LLMs) is aggregated through noticeable patterns for long
context processing. Our observations reveal that LLMs aggregate information
through Pyramidal Information Funneling where attention is scattering widely in
lower layers, progressively consolidating within specific contexts, and
ultimately focusin on critical tokens (a.k.a massive activation or attention
sink) in higher layers. Motivated by these insights, we developed PyramidKV, a
novel and effective KV cache compression method. This approach dynamically
adjusts the KV cache size across different layers, allocating more cache in
lower layers and less in higher ones, diverging from traditional methods that
maintain a uniform KV cache size. Our experimental evaluations, utilizing the
LongBench benchmark, show that PyramidKV matches the performance of models with
a full KV cache while retaining only 12% of the KV cache, thus significantly
reducing memory usage. In scenarios emphasizing memory efficiency, where only
0.7% of the KV cache is maintained, PyramidKV surpasses other KV cache
compression techniques achieving up to a 20.5 absolute accuracy improvement on
TREC. | cs.CL | PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling |
2024-06-04T07:43:33Z | http://arxiv.org/pdf/2406.02061v2 | http://arxiv.org/abs/2406.02061v2 | Marianna Nezhurina, Lucia Cipolina-Kun, Mehdi Cherti, Jenia Jitsev | 2024-06-05T23:23:54Z | cs.LG, cs.AI, cs.CL | Large Language Models (LLMs) are often described as being instances of
foundation models - that is, models that transfer strongly across various tasks
and conditions in few-show or zero-shot manner, while exhibiting scaling laws
that predict function improvement when increasing the pre-training scale. These
claims of excelling in different functions and tasks rely on measurements taken
across various sets of standardized benchmarks showing high scores for such
models. We demonstrate here a dramatic breakdown of function and reasoning
capabilities of state-of-the-art models trained at the largest available scales
which claim strong function, using a simple, short, conventional common sense
problem formulated in concise natural language, easily solvable by humans. The
breakdown is dramatic, as models also express strong overconfidence in their
wrong solutions, while providing often non-sensical "reasoning"-like
explanations akin to confabulations to justify and backup the validity of their
clearly failed responses, making them sound plausible. Various standard
interventions in an attempt to get the right solution, like various type of
enhanced prompting, or urging the models to reconsider the wrong solutions
again by multi step re-evaluation, fail. We take these initial observations to
the scientific and technological community to stimulate urgent re-assessment of
the claimed capabilities of current generation of LLMs, Such re-assessment also
requires common action to create standardized benchmarks that would allow
proper detection of such basic reasoning deficits that obviously manage to
remain undiscovered by current state-of-the-art evaluation procedures and
benchmarks. Code for reproducing experiments in the paper and raw experiments
data can be found at https://github.com/LAION-AI/AIW | cs.LG | Alice in Wonderland: Simple Tasks Showing Complete Reasoning Breakdown in State-Of-the-Art Large Language Models |
2024-06-04T07:43:12Z | http://arxiv.org/pdf/2406.02060v1 | http://arxiv.org/abs/2406.02060v1 | Valeriya Goloviznina, Evgeny Kotelnikov | 2024-06-04T07:43:12Z | cs.CL, cs.AI | Interpretability and explainability of AI are becoming increasingly important
in light of the rapid development of large language models (LLMs). This paper
investigates the interpretation of LLMs in the context of the knowledge-based
question answering. The main hypothesis of the study is that correct and
incorrect model behavior can be distinguished at the level of hidden states.
The quantized models LLaMA-2-7B-Chat, Mistral-7B, Vicuna-7B and the MuSeRC
question-answering dataset are used to test this hypothesis. The results of the
analysis support the proposed hypothesis. We also identify the layers which
have a negative effect on the model's behavior. As a prospect of practical
application of the hypothesis, we propose to train such "weak" layers
additionally in order to improve the quality of the task solution. | cs.CL | I've got the "Answer"! Interpretation of LLMs Hidden States in Question Answering |
2024-06-04T07:31:06Z | http://arxiv.org/pdf/2406.02050v2 | http://arxiv.org/abs/2406.02050v2 | Hitomi Yanaka, Namgi Han, Ryoma Kumon, Jie Lu, Masashi Takeshita, Ryo Sekizawa, Taisei Kato, Hiromi Arai | 2024-06-05T07:56:11Z | cs.CL | With the development of Large Language Models (LLMs), social biases in the
LLMs have become a crucial issue. While various benchmarks for social biases
have been provided across languages, the extent to which Japanese LLMs exhibit
social biases has not been fully investigated. In this study, we construct the
Japanese Bias Benchmark dataset for Question Answering (JBBQ) based on the
English bias benchmark BBQ, and analyze social biases in Japanese LLMs. The
results show that while current Japanese LLMs improve their accuracies on JBBQ
by instruction-tuning, their bias scores become larger. In addition, augmenting
their prompts with warning about social biases reduces the effect of biases in
some models. | cs.CL | Analyzing Social Biases in Japanese Large Language Models |
2024-06-04T07:27:36Z | http://arxiv.org/pdf/2406.02044v1 | http://arxiv.org/abs/2406.02044v1 | Hussein Jawad, Nicolas J. -B. BRUNEL | 2024-06-04T07:27:36Z | cs.CL, cs.LG | Large Language Models (LLMs) have surged in popularity in recent months, yet
they possess concerning capabilities for generating harmful content when
manipulated. This study introduces the Query-Response Optimization Attack
(QROA), an optimization-based strategy designed to exploit LLMs through a
black-box, query-only interaction. QROA adds an optimized trigger to a
malicious instruction to compel the LLM to generate harmful content. Unlike
previous approaches, QROA does not require access to the model's logit
information or any other internal data and operates solely through the standard
query-response interface of LLMs. Inspired by deep Q-learning and Greedy
coordinate descent, the method iteratively updates tokens to maximize a
designed reward function. We tested our method on various LLMs such as Vicuna,
Falcon, and Mistral, achieving an Attack Success Rate (ASR) over 80\%. We also
tested the model against Llama2-chat, the fine-tuned version of Llama2 designed
to resist Jailbreak attacks, achieving good ASR with a suboptimal initial
trigger seed. This study demonstrates the feasibility of generating jailbreak
attacks against deployed LLMs in the public domain using black-box optimization
methods, enabling more comprehensive safety testing of LLMs. | cs.CL | QROA: A Black-Box Query-Response Optimization Attack on LLMs |
2024-06-04T07:13:23Z | http://arxiv.org/pdf/2406.02030v2 | http://arxiv.org/abs/2406.02030v2 | Junlin Lee, Yequan Wang, Jing Li, Min Zhang | 2024-06-05T03:28:01Z | cs.CL, cs.AI | Multimodal reasoning with large language models (LLMs) often suffers from
hallucinations and the presence of deficient or outdated knowledge within LLMs.
Some approaches have sought to mitigate these issues by employing textual
knowledge graphs, but their singular modality of knowledge limits comprehensive
cross-modal understanding. In this paper, we propose the Multimodal Reasoning
with Multimodal Knowledge Graph (MR-MKG) method, which leverages multimodal
knowledge graphs (MMKGs) to learn rich and semantic knowledge across
modalities, significantly enhancing the multimodal reasoning capabilities of
LLMs. In particular, a relation graph attention network is utilized for
encoding MMKGs and a cross-modal alignment module is designed for optimizing
image-text alignment. A MMKG-grounded dataset is constructed to equip LLMs with
initial expertise in multimodal reasoning through pretraining. Remarkably,
MR-MKG achieves superior performance while training on only a small fraction of
parameters, approximately 2.25% of the LLM's parameter size. Experimental
results on multimodal question answering and multimodal analogy reasoning tasks
demonstrate that our MR-MKG method outperforms previous state-of-the-art
models. | cs.CL | Multimodal Reasoning with Multimodal Knowledge Graph |
2024-06-04T06:57:47Z | http://arxiv.org/pdf/2406.02018v1 | http://arxiv.org/abs/2406.02018v1 | Manasi Sharma, Ho Chit Siu, Rohan Paleja, Jaime D. Peña | 2024-06-04T06:57:47Z | cs.CL, cs.AI, cs.HC | The emergence of Large Language Models (LLMs) has revealed a growing need for
human-AI collaboration, especially in creative decision-making scenarios where
trust and reliance are paramount. Through human studies and model evaluations
on the open-ended News Headline Generation task from the LaMP benchmark, we
analyze how the framing and presence of explanations affect user trust and
model performance. Overall, we provide evidence that adding an explanation in
the model response to justify its reasoning significantly increases
self-reported user trust in the model when the user has the opportunity to
compare various responses. Position and faithfulness of these explanations are
also important factors. However, these gains disappear when users are shown
responses independently, suggesting that humans trust all model responses,
including deceptive ones, equitably when they are shown in isolation. Our
findings urge future research to delve deeper into the nuanced evaluation of
trust in human-machine teaming systems. | cs.CL | Why Would You Suggest That? Human Trust in Language Model Responses |
2024-06-04T06:43:34Z | http://arxiv.org/pdf/2406.02009v2 | http://arxiv.org/abs/2406.02009v2 | Kun Zhou, Shengkui Zhao, Yukun Ma, Chong Zhang, Hao Wang, Dianwen Ng, Chongjia Ni, Nguyen Trung Hieu, Jia Qi Yip, Bin Ma | 2024-06-12T01:33:56Z | eess.AS, cs.CL, cs.SD | Recent language model-based text-to-speech (TTS) frameworks demonstrate
scalability and in-context learning capabilities. However, they suffer from
robustness issues due to the accumulation of errors in speech unit predictions
during autoregressive language modeling. In this paper, we propose a phonetic
enhanced language modeling method to improve the performance of TTS models. We
leverage self-supervised representations that are phonetically rich as the
training target for the autoregressive language model. Subsequently, a
non-autoregressive model is employed to predict discrete acoustic codecs that
contain fine-grained acoustic details. The TTS model focuses solely on
linguistic modeling during autoregressive training, thereby reducing the error
propagation that occurs in non-autoregressive training. Both objective and
subjective evaluations validate the effectiveness of our proposed method. | eess.AS | Phonetic Enhanced Language Modeling for Text-to-Speech Synthesis |
2024-06-04T06:34:33Z | http://arxiv.org/pdf/2406.02004v2 | http://arxiv.org/abs/2406.02004v2 | Lun Wang, Om Thakkar, Zhong Meng, Nicole Rafidi, Rohit Prabhavalkar, Arun Narayanan | 2024-06-05T21:44:10Z | cs.CR, cs.CL, cs.SD, eess.AS | Gradient clipping plays a vital role in training large-scale automatic speech
recognition (ASR) models. It is typically applied to minibatch gradients to
prevent gradient explosion, and to the individual sample gradients to mitigate
unintended memorization. This work systematically investigates the impact of a
specific granularity of gradient clipping, namely per-core clip-ping (PCC),
across training a wide range of ASR models. We empirically demonstrate that PCC
can effectively mitigate unintended memorization in ASR models. Surprisingly,
we find that PCC positively influences ASR performance metrics, leading to
improved convergence rates and reduced word error rates. To avoid tuning the
additional hyperparameter introduced by PCC, we further propose a novel
variant, adaptive per-core clipping (APCC), for streamlined optimization. Our
findings highlight the multifaceted benefits of PCC as a strategy for robust,
privacy-forward ASR model training. | cs.CR | Efficiently Train ASR Models that Memorize Less and Perform Better with Per-core Clipping |
2024-06-04T06:33:13Z | http://arxiv.org/pdf/2406.02002v1 | http://arxiv.org/abs/2406.02002v1 | Shixuan Fan, Wei Wei, Wendi Li, Xian-Ling Mao, Wenfeng Xie, Dangyang Chen | 2024-06-04T06:33:13Z | cs.CL, cs.AI | The core of the dialogue system is to generate relevant, informative, and
human-like responses based on extensive dialogue history. Recently, dialogue
generation domain has seen mainstream adoption of large language models (LLMs),
due to its powerful capability in generating utterances. However, there is a
natural deficiency for such models, that is, inherent position bias, which may
lead them to pay more attention to the nearby utterances instead of causally
relevant ones, resulting in generating irrelevant and generic responses in
long-term dialogue. To alleviate such problem, in this paper, we propose a
novel method, named Causal Perception long-term Dialogue framework (CPD), which
employs perturbation-based causal variable discovery method to extract casually
relevant utterances from the dialogue history and enhances model causal
perception during fine-tuning. Specifically, a local-position awareness method
is proposed in CPD for inter-sentence position correlation elimination, which
helps models extract causally relevant utterances based on perturbations. Then,
a casual-perception fine-tuning strategy is also proposed, to enhance the
capability of discovering the causal invariant factors, by differently
perturbing causally relevant and non-casually relevant ones for response
generation. Experimental results on two datasets prove that our proposed method
can effectively alleviate the position bias for multiple LLMs and achieve
significant progress compared with existing baselines. | cs.CL | Position Debiasing Fine-Tuning for Causal Perception in Long-Term Dialogue |
2024-06-04T06:09:49Z | http://arxiv.org/pdf/2406.01988v1 | http://arxiv.org/abs/2406.01988v1 | Shixuan Fan, Wei Wei, Xiaofei Wen, Xianling Mao, Jixiong Chen, Dangyang Chen | 2024-06-04T06:09:49Z | cs.CL, cs.AI | Recently, the topic-grounded dialogue (TGD) system has become increasingly
popular as its powerful capability to actively guide users to accomplish
specific tasks through topic-guided conversations. Most existing works utilize
side information (\eg topics or personas) in isolation to enhance the topic
selection ability. However, due to disregarding the noise within these
auxiliary information sources and their mutual influence, current models tend
to predict user-uninteresting and contextually irrelevant topics. To build
user-engaging and coherent dialogue agent, we propose a \textbf{P}ersonalized
topic s\textbf{E}lection model for \textbf{T}opic-grounded \textbf{D}ialogue,
named \textbf{PETD}, which takes account of the interaction of side information
to selectively aggregate such information for more accurately predicting
subsequent topics. Specifically, we evaluate the correlation between global
topics and personas and selectively incorporate the global topics aligned with
user personas. Furthermore, we propose a contrastive learning based persona
selector to filter out irrelevant personas under the constraint of lacking
pertinent persona annotations. Throughout the selection and generation, diverse
relevant side information is considered. Extensive experiments demonstrate that
our proposed method can generate engaging and diverse responses, outperforming
state-of-the-art baselines across various evaluation metrics. | cs.CL | Personalized Topic Selection Model for Topic-Grounded Dialogue |
2024-06-04T05:51:43Z | http://arxiv.org/pdf/2406.01983v1 | http://arxiv.org/abs/2406.01983v1 | Bichen Wang, Yuzhe Zi, Yixin Sun, Yanyan Zhao, Bing Qin | 2024-06-04T05:51:43Z | cs.CL | With the passage of the Right to Be Forgotten (RTBF) regulations and the
scaling up of language model training datasets, research on model unlearning in
large language models (LLMs) has become more crucial. Before the era of LLMs,
machine unlearning research focused mainly on classification tasks in models
with small parameters. In these tasks, the content to be forgotten or retained
is clear and straightforward. However, as parameter sizes have grown and tasks
have become more complex, balancing forget quality and model utility has become
more challenging, especially in scenarios involving personal data instead of
classification results. Existing methods based on gradient ascent and its
variants often struggle with this balance, leading to unintended information
loss or partial forgetting. To address this challenge, we propose RKLD, a novel
\textbf{R}everse \textbf{KL}-Divergence-based Knowledge \textbf{D}istillation
unlearning algorithm for LLMs targeting the unlearning of personal information.
Through RKLD, we achieve significant forget quality and effectively maintain
the model utility in our experiments. | cs.CL | RKLD: Reverse KL-Divergence-based Knowledge Distillation for Unlearning Personal Information in Large Language Models |
2024-06-04T05:47:17Z | http://arxiv.org/pdf/2406.01981v1 | http://arxiv.org/abs/2406.01981v1 | Yury Tokpanov, Beren Millidge, Paolo Glorioso, Jonathan Pilault, Adam Ibrahim, James Whittington, Quentin Anthony | 2024-06-04T05:47:17Z | cs.CL, cs.AI | The size of large language models (LLMs) has scaled dramatically in recent
years and their computational and data requirements have surged
correspondingly. State-of-the-art language models, even at relatively smaller
sizes, typically require training on at least a trillion tokens. This rapid
advancement has eclipsed the growth of open-source datasets available for
large-scale LLM pretraining. In this paper, we introduce Zyda (Zyphra Dataset),
a dataset under a permissive license comprising 1.3 trillion tokens, assembled
by integrating several major respected open-source datasets into a single,
high-quality corpus. We apply rigorous filtering and deduplication processes,
both within and across datasets, to maintain and enhance the quality derived
from the original datasets. Our evaluations show that Zyda not only competes
favorably with other open datasets like Dolma, FineWeb, and RefinedWeb, but
also substantially improves the performance of comparable models from the
Pythia suite. Our rigorous data processing methods significantly enhance Zyda's
effectiveness, outperforming even the best of its constituent datasets when
used independently. | cs.CL | Zyda: A 1.3T Dataset for Open Language Modeling |
2024-06-04T05:22:24Z | http://arxiv.org/pdf/2406.01976v1 | http://arxiv.org/abs/2406.01976v1 | Xiao Zhang, Miao Li, Ji Wu | 2024-06-04T05:22:24Z | cs.CL | Language models can learn sophisticated language understanding skills from
fitting raw text. They also unselectively learn useless corpus statistics and
biases, especially during finetuning on domain-specific corpora. In this paper,
we propose a simple modification to causal language modeling called conditional
finetuning, which performs language modeling conditioned on a context. We show
that a context can "explain away" certain corpus statistics and make the model
avoid learning them. In this fashion, conditional finetuning achieves selective
learning from a corpus, learning knowledge useful for downstream tasks while
avoiding learning useless corpus statistics like topic biases. This selective
learning effect leads to less forgetting and better stability-plasticity
tradeoff in domain finetuning, potentially benefitting lifelong learning with
language models. | cs.CL | Conditional Language Learning with Context |
2024-06-04T04:17:40Z | http://arxiv.org/pdf/2406.06576v3 | http://arxiv.org/abs/2406.06576v3 | Owen Dugan, Donato Manuel Jimenez Beneto, Charlotte Loh, Zhuo Chen, Rumen Dangovski, Marin Soljačić | 2024-06-29T19:13:23Z | cs.CL, cs.AI, cs.LG | Despite significant advancements in text generation and reasoning, Large
Language Models (LLMs) still face challenges in accurately performing complex
arithmetic operations. To achieve accurate calculations, language model systems
often enable LLMs to generate code for arithmetic operations. However, this
approach compromises speed and security and, if finetuning is involved, risks
the language model losing prior capabilities. We propose a framework that
enables exact arithmetic in \textit{a single autoregressive step}, providing
faster, more secure, and more interpretable LLM systems with arithmetic
capabilities. We use the hidden states of an LLM to control a symbolic
architecture which performs arithmetic. Our implementation using Llama 3 8B
Instruct with OccamNet as a symbolic model (OccamLlama) achieves 100\% accuracy
on single arithmetic operations
($+,-,\times,\div,\sin{},\cos{},\log{},\exp{},\sqrt{}$), outperforming GPT 4o
and on par with GPT 4o using a code interpreter. OccamLlama also outperforms
GPT 4o both with and without a code interpreter on mathematical problem solving
benchmarks involving challenging arithmetic, thus enabling small LLMs to match
the arithmetic performance of even much larger models. We will make our code
public shortly. | cs.CL | OccamLLM: Fast and Exact Language Model Arithmetic in a Single Step |
2024-06-04T03:58:14Z | http://arxiv.org/pdf/2406.01946v2 | http://arxiv.org/abs/2406.01946v2 | Tong Zhou, Xuandong Zhao, Xiaolin Xu, Shaolei Ren | 2024-06-17T15:11:11Z | cs.CR, cs.CL | Text watermarks for large language models (LLMs) have been commonly used to
identify the origins of machine-generated content, which is promising for
assessing liability when combating deepfake or harmful content. While existing
watermarking techniques typically prioritize robustness against removal
attacks, unfortunately, they are vulnerable to spoofing attacks: malicious
actors can subtly alter the meanings of LLM-generated responses or even forge
harmful content, potentially misattributing blame to the LLM developer. To
overcome this, we introduce a bi-level signature scheme, Bileve, which embeds
fine-grained signature bits for integrity checks (mitigating spoofing attacks)
as well as a coarse-grained signal to trace text sources when the signature is
invalid (enhancing detectability) via a novel rank-based sampling strategy.
Compared to conventional watermark detectors that only output binary results,
Bileve can differentiate 5 scenarios during detection, reliably tracing text
provenance and regulating LLMs. The experiments conducted on OPT-1.3B and
LLaMA-7B demonstrate the effectiveness of Bileve in defeating spoofing attacks
with enhanced detectability. | cs.CR | Bileve: Securing Text Provenance in Large Language Models Against Spoofing with Bi-level Signature |
2024-06-04T03:54:53Z | http://arxiv.org/pdf/2406.01943v1 | http://arxiv.org/abs/2406.01943v1 | Nik Bear Brown | 2024-06-04T03:54:53Z | cs.CL, cs.AI, 2020: 68T50, 68Q25, I.2.7; F.2.2 | This paper surveys evaluation techniques to enhance the trustworthiness and
understanding of Large Language Models (LLMs). As reliance on LLMs grows,
ensuring their reliability, fairness, and transparency is crucial. We explore
algorithmic methods and metrics to assess LLM performance, identify weaknesses,
and guide development towards more trustworthy applications. Key evaluation
metrics include Perplexity Measurement, NLP metrics (BLEU, ROUGE, METEOR,
BERTScore, GLEU, Word Error Rate, Character Error Rate), Zero-Shot and Few-Shot
Learning Performance, Transfer Learning Evaluation, Adversarial Testing, and
Fairness and Bias Evaluation. We introduce innovative approaches like LLMMaps
for stratified evaluation, Benchmarking and Leaderboards for competitive
assessment, Stratified Analysis for in-depth understanding, Visualization of
Blooms Taxonomy for cognitive level accuracy distribution, Hallucination Score
for quantifying inaccuracies, Knowledge Stratification Strategy for
hierarchical analysis, and Machine Learning Models for Hierarchy Generation.
Human Evaluation is highlighted for capturing nuances that automated metrics
may miss. These techniques form a framework for evaluating LLMs, aiming to
enhance transparency, guide development, and establish user trust. Future
papers will describe metric visualization and demonstrate each approach on
practical examples. | cs.CL | Enhancing Trust in LLMs: Algorithms for Comparing and Interpreting LLMs |
2024-06-04T03:48:08Z | http://arxiv.org/pdf/2406.01940v1 | http://arxiv.org/abs/2406.01940v1 | Jianqiao Lu, Zhengying Liu, Yingjia Wan, Yinya Huang, Haiming Wang, Zhicheng Yang, Jing Tang, Zhijiang Guo | 2024-06-04T03:48:08Z | cs.CL, cs.LG, cs.LO | Autoformalization, the conversion of natural language mathematics into formal
languages, offers significant potential for advancing mathematical reasoning.
However, existing efforts are limited to formal languages with substantial
online corpora and struggle to keep pace with rapidly evolving languages like
Lean 4. To bridge this gap, we propose a new benchmark \textbf{Form}alization
for \textbf{L}ean~\textbf{4} (\textbf{\name}) designed to evaluate the
autoformalization capabilities of large language models (LLMs). This benchmark
encompasses a comprehensive assessment of questions, answers, formal
statements, and proofs. Additionally, we introduce a
\textbf{P}rocess-\textbf{S}upervised \textbf{V}erifier (\textbf{PSV}) model
that leverages the precise feedback from Lean 4 compilers to enhance
autoformalization. Our experiments demonstrate that the PSV method improves
autoformalization, enabling higher accuracy using less filtered training data.
Furthermore, when fine-tuned with data containing detailed process information,
PSV can leverage the data more effectively, leading to more significant
improvements in autoformalization for Lean 4. Our dataset and code are
available at \url{https://github.com/rookie-joe/PDA}. | cs.CL | Process-Driven Autoformalization in Lean 4 |
2024-06-04T03:35:25Z | http://arxiv.org/pdf/2406.01934v2 | http://arxiv.org/abs/2406.01934v2 | Zefeng Zhang, Jiawei Sheng, Chuang Zhang, Yunzhi Liang, Wenyuan Zhang, Siqi Wang, Tingwen Liu | 2024-06-05T12:13:56Z | cs.CL | Multimodal Entity Linking (MEL) aims to link ambiguous mentions in multimodal
contexts to entities in a multimodal knowledge graph. A pivotal challenge is to
fully leverage multi-element correlations between mentions and entities to
bridge modality gap and enable fine-grained semantic matching. Existing methods
attempt several local correlative mechanisms, relying heavily on the
automatically learned attention weights, which may over-concentrate on partial
correlations. To mitigate this issue, we formulate the correlation assignment
problem as an optimal transport (OT) problem, and propose a novel MEL
framework, namely OT-MEL, with OT-guided correlation assignment. Thereby, we
exploit the correlation between multimodal features to enhance multimodal
fusion, and the correlation between mentions and entities to enhance
fine-grained matching. To accelerate model prediction, we further leverage
knowledge distillation to transfer OT assignment knowledge to attention
mechanism. Experimental results show that our model significantly outperforms
previous state-of-the-art baselines and confirm the effectiveness of the
OT-guided correlation assignment. | cs.CL | Optimal Transport Guided Correlation Assignment for Multimodal Entity Linking |
2024-06-04T03:31:09Z | http://arxiv.org/pdf/2406.01931v2 | http://arxiv.org/abs/2406.01931v2 | Youcheng Huang, Jingkun Tang, Duanyu Feng, Zheng Zhang, Wenqiang Lei, Jiancheng Lv, Anthony G. Cohn | 2024-06-05T07:21:19Z | cs.CL | People tell lies when seeking rewards. Large language models (LLMs) are
aligned to human values with reinforcement learning where they get rewards if
they satisfy human preference. We find that this also induces dishonesty in
helpful and harmless alignment where LLMs tell lies in generating harmless
responses. Using the latest interpreting tools, we detect dishonesty, show how
LLMs can be harmful if their honesty is increased, and analyze such conflicts
at the parameter-level. Given these preliminaries and the hypothesis that
reward-seeking stimulates dishonesty, we theoretically show that the dishonesty
can in-turn decrease the alignment performances and augment reward-seeking
alignment with representation regularization. Extensive results, including
GPT-4 annotated win-rates, perplexities, and cases studies demonstrate that we
can train more honest, helpful, and harmless LLMs. We will make all our codes
and results be open-sourced upon this paper's acceptance. | cs.CL | Dishonesty in Helpful and Harmless Alignment |
2024-06-04T03:00:55Z | http://arxiv.org/pdf/2406.01919v1 | http://arxiv.org/abs/2406.01919v1 | Chenyang Huang, Abbas Ghaddar, Ivan Kobyzev, Mehdi Rezagholizadeh, Osmar R. Zaiane, Boxing Chen | 2024-06-04T03:00:55Z | cs.CL | Recently, there has been considerable attention on detecting hallucinations
and omissions in Machine Translation (MT) systems. The two dominant approaches
to tackle this task involve analyzing the MT system's internal states or
relying on the output of external tools, such as sentence similarity or MT
quality estimators. In this work, we introduce OTTAWA, a novel Optimal
Transport (OT)-based word aligner specifically designed to enhance the
detection of hallucinations and omissions in MT systems. Our approach
explicitly models the missing alignments by introducing a "null" vector, for
which we propose a novel one-side constrained OT setting to allow an adaptive
null alignment. Our approach yields competitive results compared to
state-of-the-art methods across 18 language pairs on the HalOmi benchmark. In
addition, it shows promising features, such as the ability to distinguish
between both error types and perform word-level detection without accessing the
MT system's internal states. | cs.CL | OTTAWA: Optimal TransporT Adaptive Word Aligner for Hallucination and Omission Translation Errors Detection |
2024-06-04T02:51:26Z | http://arxiv.org/pdf/2406.01914v1 | http://arxiv.org/abs/2406.01914v1 | Yu Tian, Tianqi Shao, Tsukasa Demizu, Xuyang Wu, Hsin-Tai Wu | 2024-06-04T02:51:26Z | cs.CV, cs.AI, cs.CL | Head pose estimation (HPE) task requires a sophisticated understanding of 3D
spatial relationships and precise numerical output of yaw, pitch, and roll
Euler angles. Previous HPE studies are mainly based on Non-large language
models (Non-LLMs), which rely on close-up human heads cropped from the full
image as inputs and lack robustness in real-world scenario. In this paper, we
present a novel framework to enhance the HPE prediction task by leveraging the
visual grounding capability of CogVLM. CogVLM is a vision language model (VLM)
with grounding capability of predicting object bounding boxes (BBoxes), which
enables HPE training and prediction using full image information input. To
integrate the HPE task into the VLM, we first cop with the catastrophic
forgetting problem in large language models (LLMs) by investigating the
rehearsal ratio in the data rehearsal method. Then, we propose and validate a
LoRA layer-based model merging method, which keeps the integrity of parameters,
to enhance the HPE performance in the framework. The results show our
HPE-CogVLM achieves a 31.5\% reduction in Mean Absolute Error for HPE
prediction over the current Non-LLM based state-of-the-art in cross-dataset
evaluation. Furthermore, we compare our LoRA layer-based model merging method
with LoRA fine-tuning only and other merging methods in CogVLM. The results
demonstrate our framework outperforms them in all HPE metrics. | cs.CV | HPE-CogVLM: New Head Pose Grounding Task Exploration on Vision Language Model |
2024-06-04T02:00:07Z | http://arxiv.org/pdf/2406.01895v1 | http://arxiv.org/abs/2406.01895v1 | Mahdi Sabbaghi, George Pappas, Hamed Hassani, Surbhi Goel | 2024-06-04T02:00:07Z | cs.LG, cs.CL, stat.ML | Despite the success of Transformers on language understanding, code
generation, and logical reasoning, they still fail to generalize over length on
basic arithmetic tasks such as addition and multiplication. A major reason
behind this failure is the vast difference in structure between numbers and
text; For example, the numbers are typically parsed from right to left, and
there is a correspondence between digits at the same position across different
numbers. In contrast, for text, such symmetries are quite unnatural. In this
work, we propose to encode these semantics explicitly into the model via
modified number formatting and custom positional encodings. Empirically, our
method allows a Transformer trained on numbers with at most 5-digits for
addition and multiplication to generalize up to 50-digit numbers, without using
additional data for longer sequences. We further demonstrate that traditional
absolute positional encodings (APE) fail to generalize to longer sequences,
even when trained with augmented data that captures task symmetries. To
elucidate the importance of explicitly encoding structure, we prove that
explicit incorporation of structure via positional encodings is necessary for
out-of-distribution generalization. Finally, we pinpoint other challenges
inherent to length generalization beyond capturing symmetries, in particular
complexity of the underlying task, and propose changes in the training
distribution to address them. | cs.LG | Explicitly Encoding Structural Symmetry is Key to Length Generalization in Arithmetic Tasks |
2024-06-04T01:20:14Z | http://arxiv.org/pdf/2406.01879v1 | http://arxiv.org/abs/2406.01879v1 | Haiming Wu, Hanqing Zhang, Richeng Xuan, Dawei Song | 2024-06-04T01:20:14Z | cs.CL | Chinese Spelling Check (CSC) aims to detect and correct potentially
misspelled characters in Chinese sentences. Naturally, it involves the
detection and correction subtasks, which interact with each other dynamically.
Such interactions are bi-directional, i.e., the detection result would help
reduce the risk of over-correction and under-correction while the knowledge
learnt from correction would help prevent false detection. Current CSC
approaches are of two types: correction-only or single-directional
detection-to-correction interactive frameworks. Nonetheless, they overlook the
bi-directional interactions between detection and correction. This paper aims
to fill the gap by proposing a Bi-directional Detector-Corrector framework for
CSC (Bi-DCSpell). Notably, Bi-DCSpell contains separate detection and
correction encoders, followed by a novel interactive learning module
facilitating bi-directional feature interactions between detection and
correction to improve each other's representation learning. Extensive
experimental results demonstrate a robust correction performance of Bi-DCSpell
on widely used benchmarking datasets while possessing a satisfactory detection
ability. | cs.CL | Bi-DCSpell: A Bi-directional Detector-Corrector Interactive Framework for Chinese Spelling Check |
2024-06-04T01:08:00Z | http://arxiv.org/pdf/2406.01876v1 | http://arxiv.org/abs/2406.01876v1 | Xuanqing Liu, Luyang Kong, Runhui Wang, Patrick Song, Austin Nevins, Henrik Johnson, Nimish Amlathe, Davor Golac | 2024-06-04T01:08:00Z | cs.DB, cs.AI, cs.CL, cs.IR, cs.LG | Schema matching constitutes a pivotal phase in the data ingestion process for
contemporary database systems. Its objective is to discern pairwise
similarities between two sets of attributes, each associated with a distinct
data table. This challenge emerges at the initial stages of data analytics,
such as when incorporating a third-party table into existing databases to
inform business insights. Given its significance in the realm of database
systems, schema matching has been under investigation since the 2000s. This
study revisits this foundational problem within the context of large language
models. Adhering to increasingly stringent data security policies, our focus
lies on the zero-shot and few-shot scenarios: the model should analyze only a
minimal amount of customer data to execute the matching task, contrasting with
the conventional approach of scrutinizing the entire data table. We emphasize
that the zero-shot or few-shot assumption is imperative to safeguard the
identity and privacy of customer data, even at the potential cost of accuracy.
The capability to accurately match attributes under such stringent requirements
distinguishes our work from previous literature in this domain. | cs.DB | GRAM: Generative Retrieval Augmented Matching of Data Schemas in the Context of Data Security |
2024-06-04T01:02:22Z | http://arxiv.org/pdf/2406.01873v2 | http://arxiv.org/abs/2406.01873v2 | Qian Lou, Xin Liang, Jiaqi Xue, Yancheng Zhang, Rui Xie, Mengxin Zheng | 2024-06-05T15:53:01Z | cs.CL, cs.CR, cs.LG | It is imperative to ensure the stability of every prediction made by a
language model; that is, a language's prediction should remain consistent
despite minor input variations, like word substitutions. In this paper, we
investigate the problem of certifying a language model's robustness against
Universal Text Perturbations (UTPs), which have been widely used in universal
adversarial attacks and backdoor attacks. Existing certified robustness based
on random smoothing has shown considerable promise in certifying the
input-specific text perturbations (ISTPs), operating under the assumption that
any random alteration of a sample's clean or adversarial words would negate the
impact of sample-wise perturbations. However, with UTPs, masking only the
adversarial words can eliminate the attack. A naive method is to simply
increase the masking ratio and the likelihood of masking attack tokens, but it
leads to a significant reduction in both certified accuracy and the certified
radius due to input corruption by extensive masking. To solve this challenge,
we introduce a novel approach, the superior prompt search method, designed to
identify a superior prompt that maintains higher certified accuracy under
extensive masking. Additionally, we theoretically motivate why ensembles are a
particularly suitable choice as base prompts for random smoothing. The method
is denoted by superior prompt ensembling technique. We also empirically confirm
this technique, obtaining state-of-the-art results in multiple settings. These
methodologies, for the first time, enable high certified accuracy against both
UTPs and ISTPs. The source code of CR-UTP is available at \url
{https://github.com/UCFML-Research/CR-UTP}. | cs.CL | CR-UTP: Certified Robustness against Universal Text Perturbations on Large Language Models |
2024-06-04T00:37:29Z | http://arxiv.org/pdf/2406.01866v2 | http://arxiv.org/abs/2406.01866v2 | Ashwin Rao, Nazanin Sabri, Siyi Guo, Louiqa Raschid, Kristina Lerman | 2024-06-10T20:33:51Z | cs.CL, cs.CY, cs.SI | Effective communication during health crises is critical, with social media
serving as a key platform for public health experts (PHEs) to engage with the
public. However, it also amplifies pseudo-experts promoting contrarian views.
Despite its importance, the role of emotional and moral language in PHEs'
communication during COVID-19 remains under explored. This study examines how
PHEs and pseudo-experts communicated on Twitter during the pandemic, focusing
on emotional and moral language and their engagement with political elites.
Analyzing tweets from 489 PHEs and 356 pseudo-experts from January 2020 to
January 2021, alongside public responses, we identified key priorities and
differences in messaging strategy. PHEs prioritize masking, healthcare,
education, and vaccines, using positive emotional language like optimism. In
contrast, pseudo-experts discuss therapeutics and lockdowns more frequently,
employing negative emotions like pessimism and disgust. Negative emotional and
moral language tends to drive engagement, but positive language from PHEs
fosters positivity in public responses. PHEs exhibit liberal partisanship,
expressing more positivity towards liberals and negativity towards conservative
elites, while pseudo-experts show conservative partisanship. These findings
shed light on the polarization of COVID-19 discourse and underscore the
importance of strategic use of emotional and moral language by experts to
mitigate polarization and enhance public trust. | cs.CL | #EpiTwitter: Public Health Messaging During the COVID-19 Pandemic |
2024-06-04T00:30:37Z | http://arxiv.org/pdf/2406.01863v1 | http://arxiv.org/abs/2406.01863v1 | Jiexin Wang, Adam Jatowt, Yi Cai | 2024-06-04T00:30:37Z | cs.CL | In the evolving field of Natural Language Processing, understanding the
temporal context of text is increasingly crucial. This study investigates
methods to incorporate temporal information during pre-training, aiming to
achieve effective time-aware language representation for improved performance
on time-related tasks. In contrast to common pre-trained models like BERT,
which rely on synchronic document collections such as BookCorpus and Wikipedia,
our research introduces BiTimeBERT 2.0, a novel language model pre-trained on a
temporal news article collection. BiTimeBERT 2.0 utilizes this temporal news
collection, focusing on three innovative pre-training objectives: Time-Aware
Masked Language Modeling (TAMLM), Document Dating (DD), and Time-Sensitive
Entity Replacement (TSER). Each objective targets a unique aspect of temporal
information. TAMLM is designed to enhance the understanding of temporal
contexts and relations, DD integrates document timestamps as chronological
markers, and TSER focuses on the temporal dynamics of "Person" entities,
recognizing their inherent temporal significance. The experimental results
consistently demonstrate that BiTimeBERT 2.0 outperforms models like BERT and
other existing pre-trained models, achieving substantial gains across a variety
of downstream NLP tasks and applications where time plays a pivotal role. | cs.CL | Towards Effective Time-Aware Language Representation: Exploring Enhanced Temporal Understanding in Language Models |
2024-06-04T00:09:43Z | http://arxiv.org/pdf/2406.01860v1 | http://arxiv.org/abs/2406.01860v1 | Jian-Qiao Zhu, Thomas L. Griffiths | 2024-06-04T00:09:43Z | cs.CL | As Large Language Models (LLMs) are increasingly deployed in real-world
settings, understanding the knowledge they implicitly use when making decisions
is critical. One way to capture this knowledge is in the form of Bayesian prior
distributions. We develop a prompt-based workflow for eliciting prior
distributions from LLMs. Our approach is based on iterated learning, a Markov
chain Monte Carlo method in which successive inferences are chained in a way
that supports sampling from the prior distribution. We validated our method in
settings where iterated learning has previously been used to estimate the
priors of human participants -- causal learning, proportion estimation, and
predicting everyday quantities. We found that priors elicited from GPT-4
qualitatively align with human priors in these settings. We then used the same
method to elicit priors from GPT-4 for a variety of speculative events, such as
the timing of the development of superhuman AI. | cs.CL | Eliciting the Priors of Large Language Models using Iterated In-Context Learning |
2024-06-04T00:01:35Z | http://arxiv.org/pdf/2406.01855v1 | http://arxiv.org/abs/2406.01855v1 | Aisha Khatun, Daniel G. Brown | 2024-06-04T00:01:35Z | cs.CL, cs.AI | Large Language Model (LLM) evaluation is currently one of the most important
areas of research, with existing benchmarks proving to be insufficient and not
completely representative of LLMs' various capabilities. We present a curated
collection of challenging statements on sensitive topics for LLM benchmarking
called TruthEval. These statements were curated by hand and contain known truth
values. The categories were chosen to distinguish LLMs' abilities from their
stochastic nature. We perform some initial analyses using this dataset and find
several instances of LLMs failing in simple tasks showing their inability to
understand simple questions. | cs.CL | TruthEval: A Dataset to Evaluate LLM Truthfulness and Reliability |
2024-06-03T23:07:18Z | http://arxiv.org/pdf/2406.01835v1 | http://arxiv.org/abs/2406.01835v1 | Mykola Trokhymovych, Indira Sen, Martin Gerlach | 2024-06-03T23:07:18Z | cs.CL, cs.AI | With over 60M articles, Wikipedia has become the largest platform for open
and freely accessible knowledge. While it has more than 15B monthly visits, its
content is believed to be inaccessible to many readers due to the lack of
readability of its text. However, previous investigations of the readability of
Wikipedia have been restricted to English only, and there are currently no
systems supporting the automatic readability assessment of the 300+ languages
in Wikipedia. To bridge this gap, we develop a multilingual model to score the
readability of Wikipedia articles. To train and evaluate this model, we create
a novel multilingual dataset spanning 14 languages, by matching articles from
Wikipedia to simplified Wikipedia and online children encyclopedias. We show
that our model performs well in a zero-shot scenario, yielding a ranking
accuracy of more than 80% across 14 languages and improving upon previous
benchmarks. These results demonstrate the applicability of the model at scale
for languages in which there is no ground-truth data available for model
fine-tuning. Furthermore, we provide the first overview on the state of
readability in Wikipedia beyond English. | cs.CL | An Open Multilingual System for Scoring Readability of Wikipedia |
2024-06-03T21:55:07Z | http://arxiv.org/pdf/2406.01806v1 | http://arxiv.org/abs/2406.01806v1 | Zhen Lin, Shubhendu Trivedi, Jimeng Sun | 2024-06-03T21:55:07Z | cs.CL, cs.AI | The advent of large language models (LLMs) has dramatically advanced the
state-of-the-art in numerous natural language generation tasks. For LLMs to be
applied reliably, it is essential to have an accurate measure of their
confidence. Currently, the most commonly used confidence score function is the
likelihood of the generated sequence, which, however, conflates semantic and
syntactic components. For instance, in question-answering (QA) tasks, an
awkward phrasing of the correct answer might result in a lower probability
prediction. Additionally, different tokens should be weighted differently
depending on the context. In this work, we propose enhancing the predicted
sequence probability by assigning different weights to various tokens using
attention values elicited from the base LLM. By employing a validation set, we
can identify the relevant attention heads, thereby significantly improving the
reliability of the vanilla sequence probability confidence measure. We refer to
this new score as the Contextualized Sequence Likelihood (CSL). CSL is easy to
implement, fast to compute, and offers considerable potential for further
improvement with task-specific prompts. Across several QA datasets and a
diverse array of LLMs, CSL has demonstrated significantly higher reliability
than state-of-the-art baselines in predicting generation quality, as measured
by the AUROC or AUARC. | cs.CL | Contextualized Sequence Likelihood: Enhanced Confidence Scores for Natural Language Generation |
2024-06-03T21:13:02Z | http://arxiv.org/pdf/2406.01789v1 | http://arxiv.org/abs/2406.01789v1 | Mario Truss, Stephan Boehm | 2024-06-03T21:13:02Z | cs.LG, cs.AI, cs.CL, cs.HC, I.2; I.2.7; K.6 | Automation of support ticket classification is crucial to improve customer
support performance and shortening resolution time for customer inquiries. This
research aims to test the applicability of automated machine learning (AutoML)
as a technology to train a machine learning model (ML model) that can classify
support tickets. The model evaluation conducted in this research shows that
AutoML can be used to train ML models with good classification performance.
Moreover, this paper fills a research gap by providing new insights into
developing AI solutions without a dedicated professional by utilizing AutoML,
which makes this technology more accessible for companies without specialized
AI departments and staff. | cs.LG | AI-based Classification of Customer Support Tickets: State of the Art and Implementation with AutoML |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.