Search Results for author:
Found 336 papers, 164 papers with code
Towards Adversarially Robust Text Classifiers by Learning to Reweight Clean Examples
Most of the existing defense methods improve the adversarial robustness by making the models adapt to the training set augmented with some adversarial examples.
Making Parameter-efficient Tuning More Efficient: A Unified Framework for Classification Tasks
Specifically, we re-formulate both token and sentence classification tasks into a unified language modeling task, and map label spaces of different tasks into the same vocabulary space.
PathQG: Neural Question Generation from Facts
Human evaluation also proves that our model is able to generate relevant and informative questions.
A Knowledge-Aware Sequence-to-Tree Network for Math Word Problem Solving
With the advancements in natural language processing tasks, math word problem solving has received increasing attention.
A Progressive Framework for Role-Aware Rumor Resolution
Existing works on rumor resolution have shown great potential in recognizing word appearance and user participation.
PlugAT: A Plug and Play Module to Defend against Textual Adversarial Attack
To reduce the potential side effects of using defense modules, we further propose a novel forgetting restricted adversarial training, which filters out bad adversarial examples that impair the performance of original ones.
A Structure-Aware Argument Encoder for Literature Discourse Analysis
Existing research for argument representation learning mainly treats tokens in the sentence equally and ignores the implied structure information of argumentative context.
Improving Abstractive Dialogue Summarization with Speaker-Aware Supervised Contrastive Learning
For dialogue summarization, the subdomain of text summarization, utterances are concatenated to flat text before being processed.
Multi-task Learning with Gradient Communication
In this paper, we describe a general framework to systematically analyze current neural models for multi-task learning, in which we find that existing models expect to disentangle features into different spaces while features learned in practice are still entangled in shared space, leaving potential hazards for other training or unseen tasks.
CQG: A Simple and Effective Controlled Generation Framework for Multi-hop Question Generation
CQG employs a simple method to generate the multi-hop questions that contain key entities in multi-hop reasoning chains, which ensure the complexity and quality of the questions.
Flooding-X: Improving BERT’s Resistance to Adversarial Attacks via Loss-Restricted Fine-Tuning
Adversarial robustness has attracted much attention recently, and the mainstream solution is adversarial training.
LFKQG: A Controlled Generation Framework with Local Fine-tuning for Question Generation over Knowledge Bases
Existing KBQG models still face two main challenges: (1) Most models often focus on the most relevant part of the answer entity, while neglecting the rest of the subgraph.
Effective Length Extrapolation via Dimension-Wise Positional Embeddings Manipulation
Instead of manipulating all dimensions equally, DPE detects the effective length for every dimension and finds the key dimensions for context extension.
Improving RL Exploration for LLM Reasoning through Retrospective Replay
We observe that for complex problems, during the early stages of training, the model exhibits strong exploratory capabilities and can identify promising solution ideas.
SocioVerse: A World Model for Social Simulation Powered by LLM Agents and A Pool of 10 Million Real-World Users
Social simulation is transforming traditional social science research by modeling human behavior through interactions between virtual individuals and their environments.
Revisiting LLM Evaluation through Mechanism Interpretability: a New Metric and Model Utility Law
In this paper, we analyze the core limitations of traditional evaluation pipelines and propose a novel metric, the Model Utilization Index (MUI), which introduces mechanism interpretability techniques to complement traditional performance metrics.
FamilyTool: A Multi-hop Personalized Tool Use Benchmark
To bridge this gap, we introduce FamilyTool, a novel benchmark grounded in a family-based knowledge graph (KG) that simulates personalized, multi-hop tool use scenarios.
Mitigating Object Hallucinations in MLLMs via Multi-Frequency Perturbations
Recently, multimodal large language models (MLLMs) have demonstrated remarkable performance in visual-language tasks.
PFDial: A Structured Dialogue Instruction Fine-tuning Method Based on UML Flowcharts
Additionally, the 8B model can surpass GPT-4o up to 43. 88% with an average of 11. 00%.
Explainable Synthetic Image Detection through Diffusion Timestep Ensembling
Based on this observation, we propose a novel detection method that amplifies these differences by progressively adding noise to the original images across multiple timesteps, and train an ensemble of classifiers on these noised images.
Better Process Supervision with Bi-directional Rewarding Signals
Process supervision, i. e., evaluating each step, is critical for complex large language model (LLM) reasoning and test-time searching with increased inference compute.
Layer-Specific Scaling of Positional Encodings for Superior Long-Context Modeling
Although large language models (LLMs) have achieved significant progress in handling long-context inputs, they still suffer from the ``lost-in-the-middle'' problem, where crucial information in the middle of the context is often underrepresented or lost.
Prior-Fitted Networks Scale to Larger Datasets When Treated as Weak Learners
Through this work, we address the challenges of efficiently handling large datasets via PFN-based models, paving the way for faster and more effective tabular data classification training and prediction process.
Thus Spake Long-Context Large Language Model
Moreover, the research on long-context LLMs has expanded from length extrapolation to a comprehensive focus on architecture, infrastructure, training, and evaluation technologies.
AutoLogi: Automated Generation of Logic Puzzles for Evaluating Reasoning Abilities of Large Language Models
While logical reasoning evaluation of Large Language Models (LLMs) has attracted significant attention, existing benchmarks predominantly rely on multiple-choice formats that are vulnerable to random guessing, leading to overestimated performance and substantial performance fluctuations.
Measuring Data Diversity for Instruction Tuning: A Systematic Analysis and A Reliable Metric
To address this, we systematically analyze 11 existing diversity measurement methods by assessing their correlation with model performance through extensive fine-tuning experiments.
How Jailbreak Defenses Work and Ensemble? A Mechanistic Investigation
Jailbreak attacks, where harmful prompts bypass generative models' built-in safety, raise serious concerns about model vulnerability.
Self-Consistency of the Internal Reward Models Improves Self-Rewarding Language Models
Recent advancements in Self-Rewarding Language Models suggest that an LLM can use its internal reward models (such as LLM-as-a-Judge) \cite{yuanself} to generate preference data, improving alignment performance without costly human annotation.
Multi-Agent Simulator Drives Language Models for Legal Intensive Interaction
Large Language Models (LLMs) have significantly advanced legal intelligence, but the scarcity of scenario data impedes the progress toward interactive legal scenarios.
Toward Relative Positional Encoding in Spiking Transformers
We provide comprehensive proof of the method's effectiveness in partially capturing relative positional information for sequential tasks.
Error Classification of Large Language Models on Math Word Problems: A Dynamically Adaptive Framework
To reduce human bias and enable fine-grained analysis of error patterns, we propose a novel framework for automated dynamic error classification in mathematical reasoning.
Human-like conceptual representations emerge from language prediction
These findings establish that structured, human-like conceptual representations can naturally emerge from language prediction without real-world grounding.
Dendritic Localized Learning: Toward Biologically Plausible Algorithm
Backpropagation is the foundational algorithm for training neural networks and a key driver of deep learning's success.
ToolHop: A Query-Driven Benchmark for Evaluating Large Language Models in Multi-Hop Tool Use
Effective evaluation of multi-hop tool use is critical for analyzing the understanding, reasoning, and function-calling capabilities of large language models (LLMs).
TL-Training: A Task-Feature-Based Framework for Training Large Language Models in Tool Use
Large language models (LLMs) achieve remarkable advancements by leveraging tools to interact with external environments, a critical step toward generalized AI.
Scaling of Search and Learning: A Roadmap to Reproduce o1 from Reinforcement Learning Perspective
OpenAI o1 represents a significant milestone in Artificial Inteiligence, which achieves expert-level performances on many challanging tasks that require strong reasoning ability. OpenAI has claimed that the main techinique behinds o1 is the reinforcement learining.
EvoWiki: Evaluating LLMs on Evolving Knowledge
Knowledge utilization is a critical aspect of LLMs, and understanding how they adapt to evolving knowledge is essential for their effective deployment.
COSEE: Consistency-Oriented Signal-Based Early Exiting via Calibrated Sample Weighting Mechanism
Early exiting is an effective paradigm for improving the inference efficiency of pre-trained language models (PLMs) by dynamically adjusting the number of executed layers for each sample.
Activating Distributed Visual Region within LLMs for Efficient and Effective Vision-Language Training and Inference
Large Vision-Language Models (LVLMs) typically learn visual capacity through visual instruction tuning, involving updates to both a projector and their LLM backbones.
From Individual to Society: A Survey on Social Simulation Driven by Large Language Model-based Agents
We categorize the simulations into three types: (1) Individual Simulation, which mimics specific individuals or demographic groups; (2) Scenario Simulation, where multiple agents collaborate to achieve goals within specific contexts; and (3) Society Simulation, which models interactions within agent societies to reflect the complexity and variety of real-world dynamics.
VLSBench: Unveiling Visual Leakage in Multimodal Safety
To explain such a counter-intuitive phenomenon, we discover a visual safety information leakage (VSIL) problem in existing multimodal safety benchmarks, i. e., the potentially risky and sensitive content in the image has been revealed in the textual query.
ObjectRelator: Enabling Cross-View Object Relation Understanding in Ego-Centric and Exo-Centric Videos
In this paper, we focus on the Ego-Exo Object Correspondence task, an emerging challenge in the field of computer vision that aims to map objects across ego-centric and exo-centric views.
Enhancing LLM Reasoning via Critique Models with Test-Time and Training-Time Supervision
Experiments show that the method improves the actor's exploration efficiency and solution diversity, especially on challenging queries, leading to a stronger reasoning model.
PIORS: Personalized Intelligent Outpatient Reception based on Large Language Model with Multi-Agents Medical Scenario Simulation
This system integrates an LLM-based reception nurse and a collaboration between LLM and hospital information system (HIS) into real outpatient reception setting, aiming to deliver personalized, high-quality, and efficient reception services.
LongSafetyBench: Long-Context LLMs Struggle with Safety Issues
With the development of large language models (LLMs), the sequence length of these models continues to increase, drawing significant attention to long-context language models.
Mitigating Tail Narrowing in LLM Self-Improvement via Socratic-Guided Sampling
Self-improvement methods enable large language models (LLMs) to generate solutions themselves and iteratively train on filtered, high-quality rationales.
Multi-Programming Language Sandbox for LLMs
We introduce MPLSandbox, an out-of-the-box multi-programming language sandbox designed to provide unified and comprehensive feedback from compiler and analysis tools for Large Language Models (LLMs).
ElectionSim: Massive Population Election Simulation Powered by Large Language Model Driven Agents
We also introduce PPE, a poll-based presidential election benchmark to assess the performance of our framework under the U. S. presidential election scenario.
Llama Scope: Extracting Millions of Features from Llama-3.1-8B with Sparse Autoencoders
We introduce a suite of 256 SAEs, trained on each layer and sublayer of the Llama-3. 1-8B-Base model, with 32K and 128K features.
AgentSense: Benchmarking Social Intelligence of Language Agents through Interactive Scenarios
To this end, we introduce AgentSense: Benchmarking Social Intelligence of Language Agents through Interactive Scenarios.
Distill Visual Chart Reasoning Ability from LLMs to MLLMs
Specifically, we employ text-based synthesizing techniques to construct chart-plotting code and produce ReachQA, a dataset containing 3k reasoning-intensive charts and 20k Q&A pairs to enhance both recognition and reasoning abilities.
Unveiling and Consulting Core Experts in Retrieval-Augmented MoE-based LLMs
In this paper, we aim to investigate these internal mechanisms within the popular Mixture-of-Expert (MoE)-based LLMs and demonstrate how to improve RAG by examining expert activations in these LLMs.
Have the VLMs Lost Confidence? A Study of Sycophancy in VLMs
Our findings indicate that the ability to prevent sycophancy is predominantly observed in higher layers of the model.
RMB: Comprehensively Benchmarking Reward Models in LLM Alignment
However, the current evaluation of RMs may not directly correspond to their alignment performance due to the limited distribution of evaluation data and evaluation methods that are not closely related to alignment objectives.
AI-Press: A Multi-Agent News Generating and Feedback Simulation System Powered by Large Language Models
The rise of various social platforms has transformed journalism.
Tell Me What You Don't Know: Enhancing Refusal Capabilities of Role-Playing Agents via Representation Space Analysis and Editing
To investigate RPAs' performance when faced with different types of conflicting requests, we develop an evaluation benchmark that includes contextual knowledge conflicting requests, parametric knowledge conflicting requests, and non-conflicting requests to assess RPAs' ability to identify conflicts and refuse to answer appropriately without over-refusing.
60 Data Points are Sufficient to Fine-Tune LLMs for Question-Answering
Large language models (LLMs) encode extensive world knowledge through pre-training on massive datasets, which can then be fine-tuned for the question-answering (QA) task.
DetectiveQA: Evaluating Long-Context Reasoning on Detective Novels
DetectiveQA focuses on evaluating the long-context reasoning ability of LLMs, which not only requires a full understanding of context but also requires extracting important evidences from the context and reasoning according to extracted evidences to answer the given questions.
Inverse-Q*: Token Level Reinforcement Learning for Aligning Large Language Models Without Preference Data
Reinforcement Learning from Human Feedback (RLHF) has proven effective in aligning large language models with human intentions, yet it often relies on complex methodologies like Proximal Policy Optimization (PPO) that require extensive hyper-parameter tuning and present challenges in sample efficiency and stability.
TransferTOD: A Generalizable Chinese Multi-Domain Task-Oriented Dialogue System with Transfer Capabilities
Task-oriented dialogue (TOD) systems aim to efficiently handle task-oriented conversations, including information collection.
Identity-Driven Hierarchical Role-Playing Agents
Utilizing large language models (LLMs) to achieve role-playing has gained great attention recently.
Overview of AI-Debater 2023: The Challenges of Argument Generation Tasks
In this paper we present the results of the AI-Debater 2023 Challenge held by the Chinese Conference on Affect Computing (CCAC 2023), and introduce the related datasets.
Case2Code: Learning Inductive Reasoning with Synthetic Data
In this paper, we hope to focus on evaluating and teaching LLMs to conduct inductive reasoning, that is, LLMs are supposed to infer underlying rules by observing examples or sequential transformations.
Synergistic Multi-Agent Framework with Trajectory Learning for Knowledge-Intensive Tasks
Recent advancements in Large Language Models (LLMs) have led to significant breakthroughs in various natural language processing tasks.
What's Wrong with Your Code Generated by Large Language Models? An Extensive Study
The increasing development of large language models (LLMs) in code generation has drawn significant attention among researchers.
HAF-RM: A Hybrid Alignment Framework for Reward Model Training
By decoupling the reward modeling procedure and incorporating hybrid supervision, our HaF-RM framework offers a principled and effective approach to enhancing the performance and alignment of reward models, a critical component in the responsible development of powerful language models.
Searching for Best Practices in Retrieval-Augmented Generation
Retrieval-augmented generation (RAG) techniques have proven to be effective in integrating up-to-date information, mitigating hallucinations, and enhancing response quality, particularly in specialized domains.
Enhancing the Capability and Robustness of Large Language Models through Reinforcement Learning-Driven Query Refinement
This strategy improves the quality of the queries, empowering LLMs to generate more truthful, benign and useful responses.
LLMs-as-Instructors: Learning from Errors Toward Automating Model Improvement
Inspired by the theory of "Learning from Errors", this framework employs an instructor LLM to meticulously analyze the specific errors within a target model, facilitating targeted and efficient training cycles.
SafeAligner: Safety Alignment against Jailbreak Attacks via Response Disparity Guidance
As the development of large language models (LLMs) rapidly advances, securing these models effectively without compromising their utility has become a pivotal area of research.
MedBench: A Comprehensive, Standardized, and Reliable Benchmarking System for Evaluating Chinese Medical Large Language Models
In this work, we introduce "MedBench", a comprehensive, standardized, and reliable benchmarking system for Chinese medical LLM.
Scaling Laws for Fact Memorization of Large Language Models
We find that LLMs' fact knowledge capacity has a linear and negative exponential law relationship with model size and training epochs, respectively.
Cross-Modality Safety Alignment
As Artificial General Intelligence (AGI) becomes increasingly integrated into various facets of human life, ensuring the safety and ethical alignment of such systems is paramount.
Inference-Time Decontamination: Reusing Leaked Benchmarks for Large Language Model Evaluation
On MMLU, using Inference-time Decontamination can lead to a decrease in the results of Phi3 and Mistral by 6. 7% and 3. 6% respectively.
Overview of the CAIL 2023 Argument Mining Track
We give a detailed overview of the CAIL 2023 Argument Mining Track, one of the Chinese AI and Law Challenge (CAIL) 2023 tracks.
SPA-VL: A Comprehensive Safety Preference Alignment Dataset for Vision Language Model
To address these limitations, we propose a Safety Preference Alignment dataset for Vision Language Models named SPA-VL.
Beyond Boundaries: Learning a Universal Entity Taxonomy across Datasets and Languages for Open Named Entity Recognition
First, we detect inconsistent entity definitions across datasets and clarify them by distinguishable label names to construct a universal taxonomy of 400+ entity types.
Promoting Data and Model Privacy in Federated Learning through Quantized LoRA
Conventional federated learning primarily aims to secure the privacy of data distributed across multiple edge devices, with the global model dispatched to edge devices for parameter updates during the learning process.
Toward Optimal LLM Alignments Using Two-Player Games
We theoretically demonstrate that this iterative reinforcement learning optimization converges to a Nash Equilibrium for the game induced by the agents.
EmbSpatial-Bench: Benchmarking Spatial Understanding for Embodied Tasks with Large Vision-Language Models
The recent rapid development of Large Vision-Language Models (LVLMs) has indicated their potential for embodied tasks. However, the critical skill of spatial understanding in embodied environments has not been thoroughly evaluated, leaving the gap between current LVLMs and qualified embodied intelligence unknown.
AgentGym: Evolving Large Language Model-based Agents across Diverse Environments
Building generalist agents that can handle diverse tasks and evolve themselves across different environments is a long-term goal in the AI community.
Advancing Spiking Neural Networks for Sequential Modeling with Central Pattern Generators
Spiking neural networks (SNNs) represent a promising approach to developing artificial neural networks that are both energy-efficient and biologically plausible.
Aggregation of Reasoning: A Hierarchical Framework for Enhancing Answer Selection in Large Language Models
Recent advancements in Chain-of-Thought prompting have facilitated significant breakthroughs for Large Language Models (LLMs) in complex reasoning tasks.
Infer Induced Sentiment of Comment Response to Video: A New Task, Dataset and Baseline
In light of this, we introduces a novel research task, Multi-modal Sentiment Analysis for Comment Response of Video Induced(MSA-CRVI), aims to inferring opinions and emotions according to the comments response to micro video.
MetaRM: Shifted Distributions Alignment via Meta-Learning
These two issues can be united as a challenge posed by the shifted distribution of the environment.
Length Generalization of Causal Transformers without Position Encoding
Generalizing to longer sentences is important for recent Transformer-based language models.
DELAN: Dual-Level Alignment for Vision-and-Language Navigation by Cross-Modal Contrastive Learning
For task completion, the agent needs to align and integrate various navigation modalities, including instruction, observation and navigation history.
Self-Demos: Eliciting Out-of-Demonstration Generalizability in Large Language Models
The generated demos strategically interpolate between existing demos and the given query, transforming the query from OOD to ID.
Subspace Defense: Discarding Adversarial Perturbations by Learning a Subspace for Clean Signals
We first empirically show that the features of either clean signals or adversarial perturbations are redundant and span in low-dimensional linear subspaces respectively with minimal overlap, and the classical low-dimensional subspace projection can suppress perturbation features out of the subspace of clean signals.
EasyJailbreak: A Unified Framework for Jailbreaking Large Language Models
This paper introduces EasyJailbreak, a unified framework simplifying the construction and evaluation of jailbreak attacks against LLMs.
Decoding Continuous Character-based Language from Non-invasive Brain Recordings
Deciphering natural language from brain activity through non-invasive devices remains a formidable challenge.
Debatrix: Multi-dimensional Debate Judge with Iterative Chronological Analysis Based on LLM
How can we construct an automated debate judge to evaluate an extensive, vibrant, multi-turn debate?
ALaRM: Align Language Models via Hierarchical Rewards Modeling
We introduce ALaRM, the first framework modeling hierarchical rewards in reinforcement learning from human feedback (RLHF), which is designed to enhance the alignment of large language models (LLMs) with human preferences.
Unveiling the Truth and Facilitating Change: Towards Agent-based Large-scale Social Movement Simulation
Simulating the response of the public and forecasting the potential impact has become increasingly important.
CodeChameleon: Personalized Encryption Framework for Jailbreaking Large Language Models
Adversarial misuse, particularly through `jailbreaking' that circumvents a model's safety and ethical protocols, poses a significant challenge for Large Language Models (LLMs).
RoCoIns: Enhancing Robustness of Large Language Models through Code-Style Instructions
In this paper, drawing inspiration from recent works that LLMs are sensitive to the design of the instructions, we utilize instructions in code style, which are more structural and less ambiguous, to replace typically natural language instructions.
Advancing Parameter Efficiency in Fine-tuning via Representation Editing
Parameter Efficient Fine-Tuning (PEFT) techniques have drawn significant attention due to their ability to yield competitive results while updating only a small portion of the adjustable parameters.
LLM-DA: Data Augmentation via Large Language Models for Few-Shot Named Entity Recognition
To overcome the limitations of existing data augmentation methods that compromise semantic integrity and address the uncertainty inherent in LLM-generated text, we leverage the distinctive characteristics of the NER task by augmenting the original data at both the contextual and entity levels.
Domain Generalization via Causal Adjustment for Cross-Domain Sentiment Analysis
Domain adaption has been widely adapted for cross-domain sentiment analysis to transfer knowledge from the source domain to the target domain.
Unveiling Linguistic Regions in Large Language Models
Furthermore, this core region exhibits significant dimensional dependence, perturbations to even a single parameter on specific dimensions leading to a loss of linguistic competence.
On the Tip of the Tongue: Analyzing Conceptual Representation in Large Language Models with Reverse-Dictionary Probe
Here we re-purpose the reverse dictionary task as a case study to probe LLMs' capacity for conceptual inference.
SoMeLVLM: A Large Vision Language Model for Social Media Processing
The powerful Large Vision Language Models make it possible to handle a variety of tasks simultaneously, but even with carefully designed prompting methods, the general domain models often fall short in aligning with the unique speaking style and context of social media tasks.
Automating Dataset Updates Towards Reliable and Timely Evaluation of Large Language Models
There are two updating strategies: 1) mimicking strategy to generate similar samples based on original data, preserving stylistic and contextual essence, and 2) extending strategy that further expands existing samples at varying cognitive levels by adapting Bloom's taxonomy of educational objectives.
LongAgent: Scaling Language Models to 128k Context through Multi-Agent Collaboration
Large language models (LLMs) have demonstrated impressive performance in understanding language and executing complex reasoning tasks.
Advancing Translation Preference Modeling with RLHF: A Step Towards Cost-Effective Solution
To address this issue, we propose a cost-effective preference learning strategy, optimizing reward models by distinguishing between human and machine translations.
Benchmark Self-Evolving: A Multi-Agent Framework for Dynamic LLM Evaluation
Towards a more scalable, robust and fine-grained evaluation, we implement six reframing operations to construct evolving instances testing LLMs against diverse queries, data noise and probing their problem-solving sub-abilities.
LLM can Achieve Self-Regulation via Hyperparameter Aware Generation
HAG extends the current paradigm in the text generation process, highlighting the feasibility of endowing the LLMs with self-regulate decoding strategies.
ToolSword: Unveiling Safety Issues of Large Language Models in Tool Learning Across Three Stages
Tool learning is widely acknowledged as a foundational approach or deploying large language models (LLMs) in real-world scenarios.
LongHeads: Multi-Head Attention is Secretly a Long Context Processor
Instead of allowing each head to attend to the full sentence, which struggles with generalizing to longer sequences due to out-of-distribution (OOD) issues, we allow each head to process in-distribution length by selecting and attending to important context chunks.
Training Large Language Models for Reasoning through Reverse Curriculum Reinforcement Learning
In this paper, we propose R$^3$: Learning Reasoning through Reverse Curriculum Reinforcement Learning (RL), a novel method that employs only outcome supervision to achieve the benefits of process supervision for large language models.
Are Large Language Models Good Prompt Optimizers?
Despite the success, the underlying mechanism of this approach remains unexplored, and the true effectiveness of LLMs as Prompt Optimizers requires further validation.
StepCoder: Improve Code Generation with Reinforcement Learning from Compiler Feedback
The advancement of large language models (LLMs) has significantly propelled the field of code generation.
Efficient and Effective Time-Series Forecasting with Spiking Neural Networks
In this paper, we propose a framework for SNNs in time-series forecasting tasks, leveraging the efficiency of spiking neurons in processing temporal information.
CauESC: A Causal Aware Model for Emotional Support Conversation
Emotional Support Conversation aims at reducing the seeker's emotional distress through supportive response.
Navigating the OverKill in Large Language Models
Large language models are meticulously aligned to be both helpful and harmless.
MouSi: Poly-Visual-Expert Vision-Language Models
This technique introduces a fusion network to unify the processing of outputs from different visual experts, while bridging the gap between image encoders and pre-trained LLMs.
Ranked #124 on
Visual Question Answering
on MM-Vet
F-Eval: Assessing Fundamental Abilities with Refined Evaluation Methods
However, these evaluation benchmarks are limited to assessing the instruction-following capabilities, overlooking the fundamental abilities that emerge during the pre-training stage.
RoTBench: A Multi-Level Benchmark for Evaluating the Robustness of Large Language Models in Tool Learning
To bridge this gap, we introduce RoTBench, a multi-level benchmark for evaluating the robustness of LLMs in tool learning.
Revisiting Jailbreaking for Large Language Models: A Representation Engineering Perspective
The recent surge in jailbreaking attacks has revealed significant vulnerabilities in Large Language Models (LLMs) when exposed to malicious inputs.
Secrets of RLHF in Large Language Models Part II: Reward Modeling
We introduce a series of novel methods to mitigate the influence of incorrect and ambiguous preferences in the dataset and fully leverage high-quality preference data.
LLaMA Beyond English: An Empirical Study on Language Capability Transfer
In recent times, substantial advancements have been witnessed in large language models (LLMs), exemplified by ChatGPT, showcasing remarkable proficiency across a range of complex tasks.
ToolEyes: Fine-Grained Evaluation for Tool Learning Capabilities of Large Language Models in Real-world Scenarios
Existing evaluations of tool learning primarily focus on validating the alignment of selected tools for large language models (LLMs) with expected outcomes.
Aligning Large Language Models with Human Preferences through Representation Engineering
Aligning large language models (LLMs) with human preferences is crucial for enhancing their utility in terms of helpfulness, truthfulness, safety, harmlessness, and interestingness.
Argue with Me Tersely: Towards Sentence-Level Counter-Argument Generation
The results show the competitiveness of our proposed framework and evaluator in counter-argument generation tasks.
A Soft Contrastive Learning-based Prompt Model for Few-shot Sentiment Analysis
Few-shot text classification has attracted great interest in both academia and industry due to the lack of labeled data in many fields.
K-ESConv: Knowledge Injection for Emotional Support Dialogue Systems via Prompt Learning
Automatic psychological counseling requires mass of professional knowledge that can be found in online counseling forums.
LoRAMoE: Alleviate World Knowledge Forgetting in Large Language Models via MoE-Style Plugin
Supervised fine-tuning (SFT) is a crucial step for large language models (LLMs), enabling them to align with human instructions and enhance their capabilities in downstream tasks.
LLMEval: A Preliminary Study on How to Evaluate Large Language Models
The three crucial questions for LLM evaluation are ``what, where, and how to evaluate''.
Exchange-of-Thought: Enhancing Large Language Model Capabilities through Cross-Model Communication
Large Language Models (LLMs) have recently made significant strides in complex reasoning tasks through the Chain-of-Thought technique.
Hi-ArG: Exploring the Integration of Hierarchical Argumentation Graphs in Language Pretraining
In this paper, we propose the Hierarchical Argumentation Graph (Hi-ArG), a new structure to organize arguments.
Making Harmful Behaviors Unlearnable for Large Language Models
Specifically, we introduce ``security vectors'', a few new parameters that can be separated from the LLM, to ensure LLM's responses are consistent with the harmful behavior.
Tailoring Personality Traits in Large Language Models via Unsupervisedly-Built Personalized Lexicons
Users can utilize UBPL to adjust the probability vectors of predicted words in the decoding phase of LLMs, thus influencing the personality expression of LLMs.
DISC-FinLLM: A Chinese Financial Large Language Model based on Multiple Experts Fine-tuning
We propose Multiple Experts Fine-tuning Framework to build a financial large language model (LLM), DISC-FinLLM.
Unveiling A Core Linguistic Region in Large Language Models
We have discovered a core region in LLMs that corresponds to linguistic competence, accounting for approximately 1% of the total model parameters.
Orthogonal Subspace Learning for Language Model Continual Learning
In this paper, we propose orthogonal low-rank adaptation (O-LoRA), a simple and efficient approach for continual learning in language models, effectively mitigating catastrophic forgetting while learning new tasks.
Are Structural Concepts Universal in Transformer Language Models? Towards Interpretable Cross-Lingual Generalization
We then propose a meta-learning-based method to learn to align conceptual spaces of different languages, which facilitates zero-shot and few-shot generalization in concept classification and also offers insights into the cross-lingual in-context learning phenomenon.
Improving Generalization of Alignment with Human Preferences through Group Invariant Learning
In this work, we propose a novel approach that can learn a consistent policy via RL across various data groups or domains.
RealBehavior: A Framework for Faithfully Characterizing Foundation Models' Human-like Behavior Mechanisms
Reports of human-like behaviors in foundation models are growing, with psychological theories providing enduring tools to investigate these behaviors.
Efficient Link Prediction via GNN Layers Induced by Negative Sampling
In contrast, \emph{edge-wise} methods rely on the formation of edge-specific subgraph embeddings to enrich the representation of pair-wise relationships, disambiguating isomorphic nodes to improve accuracy, but with increased model complexity.
RethinkingTMSC: An Empirical Study for Target-Oriented Multimodal Sentiment Classification
Recently, Target-oriented Multimodal Sentiment Classification (TMSC) has gained significant attention among scholars.
TRACE: A Comprehensive Benchmark for Continual Learning in Large Language Models
In this paper, we introduce TRACE, a novel benchmark designed to evaluate continual learning in LLMs.
SpikeCLIP: A Contrastive Language-Image Pretrained Spiking Neural Network
Spiking Neural Networks (SNNs) have emerged as a promising alternative to conventional Artificial Neural Networks (ANNs), demonstrating comparable performance in both visual and linguistic tasks while offering the advantage of improved energy efficiency.
Loose lips sink ships: Mitigating Length Bias in Reinforcement Learning from Human Feedback
Reinforcement learning from human feedback serves as a crucial bridge, aligning large language models with human and societal values.
ReForm-Eval: Evaluating Large Vision Language Models via Unified Re-Formulation of Task-Oriented Benchmarks
Recent years have witnessed remarkable progress in the development of large vision-language models (LVLMs).
DISC-LawLLM: Fine-tuning Large Language Models for Intelligent Legal Services
We propose DISC-LawLLM, an intelligent legal system utilizing large language models (LLMs) to provide a wide range of legal services.
The Rise and Potential of Large Language Model Based Agents: A Survey
Many efforts have been made to develop intelligent agents, but they mainly focus on advancement in algorithms or training strategies to enhance specific capabilities or performance on particular tasks.
SpikeBERT: A Language Spikformer Learned from BERT with Knowledge Distillation
Spiking neural networks (SNNs) offer a promising avenue to implement deep neural networks in a more energy-efficient way.
DISC-MedLLM: Bridging General Large Language Models and Real-World Medical Consultation
We propose DISC-MedLLM, a comprehensive solution that leverages Large Language Models (LLMs) to provide accurate and truthful medical response in end-to-end conversational healthcare services.
Secrets of RLHF in Large Language Models Part I: PPO
Therefore, we explore the PPO-max, an advanced version of PPO algorithm, to efficiently improve the training stability of the policy model.
On the Universal Adversarial Perturbations for Efficient Data-free Adversarial Detection
Detecting adversarial samples that are carefully crafted to fool the model is a critical step to socially-secure applications.
From Hypergraph Energy Functions to Hypergraph Neural Networks
Hypergraphs are a powerful abstraction for representing higher-order interactions between entities of interest.
Open Set Relation Extraction via Unknown-Aware Training
Inspired by text adversarial attacks, we adaptively apply small but critical perturbations to original training instances and thus synthesizing negative instances that are more likely to be mistaken by the model as known relations.
Do Large Language Models Know What They Don't Know?
Large language models (LLMs) have a wealth of knowledge that allows them to excel in various Natural Language Processing (NLP) tasks.
KNSE: A Knowledge-aware Natural Language Inference Framework for Dialogue Symptom Status Recognition
Symptom diagnosis in medical conversations aims to correctly extract both symptom entities and their status from the doctor-patient dialogue.
Self-Polish: Enhance Reasoning in Large Language Models via Problem Refinement
To enhance the multi-step reasoning capabilities of large language models, researchers have extensively explored prompting methods, notably the Chain-of-Thought (CoT) method which explicitly elicits human-like rationales.
Query Structure Modeling for Inductive Logical Reasoning Over Knowledge Graphs
The results demonstrate the effectiveness of our method on logical reasoning over KGs in both inductive and transductive settings.
A Confidence-based Partial Label Learning Model for Crowd-Annotated Named Entity Recognition
Particularly, we propose a Confidence-based Partial Label Learning (CPLL) method to integrate the prior confidence (given by annotators) and posterior confidences (learned by models) for crowd-annotated NER.
Modeling the Q-Diversity in a Min-max Play Game for Robust Optimization
Models trained with empirical risk minimization (ERM) are revealed to easily rely on spurious correlations, resulting in poor generalization.
Enhancing Contrastive Learning with Noise-Guided Attack: Towards Continual Relation Extraction in the Wild
The principle of continual relation extraction~(CRE) involves adapting to emerging novel relations while preserving od knowledge.
CodeIE: Large Code Generation Models are Better Few-Shot Information Extractors
A common practice is to recast the task into a text-to-text format such that generative LLMs of natural language (NL-LLMs) like GPT-3 can be prompted to solve it.
CausalAPM: Generalizable Literal Disentanglement for NLU Debiasing
Dataset bias, i. e., the over-reliance on dataset-specific literal heuristics, is getting increasing attention for its detrimental effect on the generalization ability of NLU models.
A Comprehensive Capability Analysis of GPT-3 and GPT-3.5 Series Models
GPT series models, such as GPT-3, CodeX, InstructGPT, ChatGPT, and so on, have gained considerable attention due to their exceptional natural language processing capabilities.
How Robust is GPT-3.5 to Predecessors? A Comprehensive Study on Language Understanding Tasks
The GPT-3. 5 models have demonstrated impressive performance in various Natural Language Processing (NLP) tasks, showcasing their strong understanding and reasoning capabilities.
Natural Language Inference
Natural Language Understanding
+1
Cross-Linguistic Syntactic Difference in Multilingual BERT: How Good is It and How Does It Affect Transfer?
We demonstrate that the distance between the distributions of different languages is highly consistent with the syntactic difference in terms of linguistic formalisms.
Rethinking Label Smoothing on Multi-hop Question Answering
Multi-Hop Question Answering (MHQA) is a significant area in question answering, requiring multiple reasoning components, including document retrieval, supporting sentence prediction, and answer span extraction.
DiffusionBERT: Improving Generative Masked Language Models with Diffusion Models
We present DiffusionBERT, a new generative masked language model based on discrete diffusion models.
Efficient Adversarial Training with Robust Early-Bird Tickets
Adversarial training is one of the most powerful methods to improve the robustness of pre-trained language models (PLMs).
Robust Lottery Tickets for Pre-trained Language Models
Recent works on Lottery Ticket Hypothesis have shown that pre-trained language models (PLMs) contain smaller matching subnetworks(winning tickets) which are capable of reaching accuracy comparable to the original models.
Late Prompt Tuning: A Late Prompt Could Be Better Than Many Prompts
Through extensive experimental results across various tasks and PTMs, we show that LPT can achieve competitive performance to full model tuning and other PETuning methods under both full-data and few-shot scenarios while possessing faster training speed and lower memory cost.
Kernel-Whitening: Overcome Dataset Bias with Isotropic Sentence Embedding
Dataset bias has attracted increasing attention recently for its detrimental effect on the generalization ability of fine-tuned models.
BERTScore is Unfair: On Social Bias in Language Model-Based Metrics for Text Generation
Automatic evaluation metrics are crucial to the development of generative systems.
Multitask Pre-training of Modular Prompt for Chinese Few-Shot Learning
MP2 is a set of combinable prompts pre-trained on 38 Chinese tasks.
COLO: A Contrastive Learning based Re-ranking Framework for One-Stage Summarization
Traditional training paradigms for extractive and abstractive summarization systems always only use token-level or sentence-level training objectives.
A Multi-Format Transfer Learning Model for Event Argument Extraction via Variational Information Bottleneck
Event argument extraction (EAE) aims to extract arguments with given roles from texts, which have been widely studied in natural language processing.
Locate Then Ask: Interpretable Stepwise Reasoning for Multi-hop Question Answering
In this paper, we propose an interpretable stepwise reasoning framework to incorporate both single-hop supporting sentence identification and single-hop question generation at each intermediate step, and utilize the inference of the current hop for the next until reasoning out the final result.
Causal Intervention Improves Implicit Sentiment Analysis
In this work, we propose a causal intervention model for Implicit Sentiment Analysis using Instrumental Variable (ISAIV).
A Unified Continuous Learning Framework for Multi-modal Knowledge Discovery and Pre-training
These two steps are iteratively performed in our framework for continuous learning.
CoNT: Contrastive Neural Text Generation
We validate CoNT on five generation tasks with ten benchmarks, including machine translation, summarization, code comment generation, data-to-text generation and commonsense generation.
What Dense Graph Do You Need for Self-Attention?
Transformers have made progress in miscellaneous tasks, but suffer from quadratic computational and memory complexities.
BBTv2: Towards a Gradient-Free Future with Large Language Models
By contrast, gradient-free methods only require the forward computation of the PTM to tune the prompt, retaining the benefits of efficient tuning and deployment.
A Benchmark for Automatic Medical Consultation System: Frameworks, Tasks and Datasets
In recent years, interest has arisen in using machine learning to improve the efficiency of automatic medical consultation and enhance patient experience.
MINER: Improving Out-of-Vocabulary Named Entity Recognition from an Information Theoretic Perspective
NER model has achieved promising performance on standard NER benchmarks.
Ranked #10 on
Named Entity Recognition (NER)
on WNUT 2017
A Simple Hash-Based Early Exiting Approach For Language Understanding and Generation
Previous works usually adopt heuristic metrics such as the entropy of internal outputs to measure instance difficulty, which suffers from generalization and threshold-tuning.
Decorrelate Irrelevant, Purify Relevant: Overcome Textual Spurious Correlations from a Feature Perspective
Most of the existing debiasing methods often identify and weaken these samples with biased features (i. e., superficial surface features that cause such spurious correlations).
MVPTR: Multi-Level Semantic Alignment for Vision-Language Pre-Training via Multi-Stage Learning
In MVPTR, we follow the nested structure of both modalities to introduce concepts as high-level semantics.
Black-Box Tuning for Language-Model-as-a-Service
In such a scenario, which we call Language-Model-as-a-Service (LMaaS), the gradients of PTMs are usually unavailable.
Plug-Tagger: A Pluggable Sequence Labeling Framework Using Language Models
Specifically, for each task, a label word set is first constructed by selecting a high-frequency word for each class respectively, and then, task-specific vectors are inserted into the inputs and optimized to manipulate the model predictions towards the corresponding label words.
Towards Efficient NLP: A Standard Evaluation and A Strong Baseline
ELUE is dedicated to depict the Pareto Frontier for various language understanding tasks, such that it can tell whether and how much a method achieves Pareto improvement.
KNN-BERT: Fine-Tuning Pre-Trained Models with KNN Classifier
Pre-trained models are widely used in fine-tuning downstream tasks with linear classifiers optimized by the cross-entropy loss, which might face robustness and stability problems.
Template-free Prompt Tuning for Few-shot NER
Prompt-based methods have been successfully applied in sentence-level few-shot learning tasks, mostly owing to the sophisticated design of templates and label words.
Constructing Phrase-level Semantic Labels to Form Multi-Grained Supervision for Image-Text Retrieval
Existing research for image text retrieval mainly relies on sentence-level supervision to distinguish matched and mismatched sentences for a query image.
Learning to Teach with Student Feedback
This one-way interaction leads to the teacher's inability to perceive the characteristics of the student and its training progress.
Fine-Grained Element Identification in Complaint Text of Internet Fraud
Existing system dealing with online complaint provides a final decision without explanations.
Defense against Synonym Substitution-based Adversarial Attacks via Dirichlet Neighborhood Ensemble
Although deep neural networks have achieved prominent performance on many NLP tasks, they are vulnerable to adversarial examples.
Align Voting Behavior with Public Statements for Legislator Representation Learning
In addition, we propose a novel task, namely hashtag usage prediction to model the ideology of legislators on Twitter.
Math Word Problem Solving with Explicit Numerical Values
In recent years, math word problem solving has received considerable attention and achieved promising results, but previous methods rarely take numerical values into consideration.
SENT: Sentence-level Distant Relation Extraction via Negative Training
In this work, we propose the use of negative training (NT), in which a model is trained using complementary labels regarding that ``the instance does not belong to these complementary labels".
