Papers by Pengyu Li

27 papers
Uncertainty Quantification for Clinical Outcome Predictions with (Large) Language Models (2025.findings-naacl)

Copied to clipboard

Challenge: Language models (LMs) have significant potential for clinical prediction tasks . however, unreliable decisions can result in significant costs due to compromised patient safety and ethical concerns .
Approach: They propose to combine ensembling and multi-tasking approaches to reduce uncertainty in EHRs by using multi-tapping methods.
Outcome: The proposed framework reduces model uncertainty in white-box and black-box settings, and improves model transparency in both settings.
UnifiedMLLM: Enabling Unified Representation for Multi-modal Multi-tasks With Large Language Model (2025.findings-naacl)

Copied to clipboard

Challenge: Representative models like LLaVA and MiniGPT-4 have great capabilities in various tasks.
Approach: They propose a unified model to represent various multi-modal tasks using a single representation.
Outcome: The proposed model outperforms existing models in a variety of tasks while maintaining generality and scalability.
Atoxia: Red-teaming Large Language Models with Target Toxic Answers (2025.findings-naacl)

Copied to clipboard

Challenge: Large language models (LLMs) are still vulnerable to generation safety vulnerabilities.
Approach: They propose a method that A**tacks LLMs with target "toxi" given a particular harmful answer, the method generates a user query and a misleading answer opening to examine the internal defects of a given LLM.
Outcome: The proposed method detects safety risks in open-source models and state-of-the-art models such as GPT-4o.
Learning to Update Natural Language Comments Based on Code Changes (2020.acl-main)

Copied to clipboard

Challenge: a novel approach to update comments based on code changes is proposed . a dataset of open-source software projects is used to train and evaluate the model .
Approach: They propose an approach that learns to correlate changes across two distinct language representations to generate a sequence of edits that are applied to the existing comment to reflect the source code modifications.
Outcome: The proposed model outperforms baselines and automatic metrics with respect to making edits.
SeqXGPT: Sentence-Level AI-Generated Text Detection (2023.emnlp-main)

Copied to clipboard

Challenge: Existing methods for sentence-level AIGT detection are weak . large language models (LLMs) can generate human-like content .
Approach: They propose a sentence-level AIGT detection challenge using LLMs as log probability lists . they propose 'check' GPT' method that uses log probability list features to detect AIGT .
Outcome: The proposed method surpasses baseline methods in sentence- and document-level detection challenges.
Sparsity-Accelerated Training for Large Language Models (2024.findings-acl)

Copied to clipboard

Challenge: Large language models (LLMs) have demonstrated proficiency across various NLP tasks but often require additional training, such as continual pre-training and supervised fine-tuning.
Approach: They propose to leverage sparsity in pre-trained LLMs to accelerate training by disregarding computations for unimportant neurons.
Outcome: The proposed framework achieves comparable or superior performance to standard training while significantly accelerating the process.
Watermarking LLMs with Weight Quantization (2023.findings-emnlp)

Copied to clipboard

Challenge: Large language models are being deployed at an astonishing speed, exposing users to high risks.
Approach: They propose a method that plants watermarks in quantization process of large language models without pre-defined triggers during inference.
Outcome: The proposed method protects model weights without pre-defined triggers . it works when the model is used in the fp32 mode and remains hidden when the models are quantized to int8 .
Adversarial Preference Optimization: Enhancing Your Alignment via RM-LLM Game (2024.findings-acl)

Copied to clipboard

Challenge: Existing methods for training large language models require additional annotations to adjust to shifted distributions.
Approach: They propose an algorithm that allows LLMs and reward models to update alternatively via a min-max game to improve their alignment.
Outcome: The proposed framework improves existing alignment baselines in terms of LLM helpfulness and harmlessness.
InferAligner: Inference-Time Alignment for Harmlessness through Cross-Model Guidance (2024.emnlp-main)

Copied to clipboard

Challenge: Existing methods for enhancing harmlessness and helpfulness of large language models (LLMs) involve complex and resource-intensive training processes.
Approach: They propose a method that decouples harmlessness from helpfulness during inference phase.
Outcome: The proposed method significantly reduces the attack success rate (ASR) of harmful instructions and jailbreak instructions while maintaining almost unchanged performance in downstream tasks.
MARCH: Multi-Agent Reinforced Check for Hallucination (2026.acl-long)

Copied to clipboard

Challenge: Existing methods to detect hallucinations suffer from inherent confirmation bias, where the verifier inadvertently reproduces the errors of the original generation.
Approach: They propose a framework that enforces rigorous factual alignment by leveraging deliberate *information asymmetry* by combining a pipeline of three specialized agents: a Solver, a Proposer, and a Checker.
Outcome: Extensive experiments across hallucination benchmarks demonstrate that MARCH substantially reduces hallucinism rates.
CARD: Cross-modal Agent Framework for Generative and Editable Residential Design (2025.emnlp-main)

Copied to clipboard

Challenge: Architectural design automation has made significant progress, but the complexity of open-world environments makes residential design a challenging task.
Approach: They propose a framework that leverages a system of specialized cross-modal agents to adapt to open-world residential design.
Outcome: The proposed framework enables users to generate and edit residential design without requiring specialized expertise.
Advancing Fine-Grained Visual Understanding with Multi-Scale Alignment in Multi-Modal Models (2025.emnlp-main)

Copied to clipboard

Challenge: Recent advances in multi-modal large language models have demonstrated remarkable capabilities in multimodal understanding, reasoning, and interaction.
Approach: They propose a method that effectively aligns and integrates multi-scale knowledge of objects . they use a pipeline that provides over 300K essential training data to enhance alignment .
Outcome: The proposed method effectively aligns and integrates multi-scale knowledge of objects, including texts, coordinates, and images.
PhysPRM: A Generative Process Reward Model with Fine-grained Diagnosis for Physics Problem Solving (2026.findings-acl)

Copied to clipboard

Challenge: Existing Large Language Models (LLMs) struggle with physics problem solving due to difficulties in decoding implicit constraints and maintaining physical consistency.
Approach: They propose a Generative PRM that treats evaluation as a generative task . it produces fine-grained diagnoses comprising critiques, final judgments, and specific error types .
Outcome: The proposed model improves performance across seven benchmarks in Best-of-N and critique refinement strategies.
Decoupled Proxy Alignment: Mitigating Language Prior Conflict for Multimodal Alignment in MLLMs (2025.findings-emnlp)

Copied to clipboard

Challenge: Recent advances in multimodal large language models focus on improving performance . however, language prior conflict leads to suboptimal vision-language alignment .
Approach: They propose a method to decouple the alignment process from language prior interference . they use a proxy LLM to detach from language interference during pretraining .
Outcome: The proposed method improves training performance and generalizes training data.
AGTAO: Robust and Stabilized LLM Unlearning via Adversarial Gating Training with Adaptive Orthogonality (2026.findings-acl)

Copied to clipboard

Challenge: Large Language Models (LLMs) unintentionally memorize sensitive data, posing privacy and security risks.
Approach: They propose a framework that reconciles unlearning efficacy and utility preservation by using a latent-space gating mechanism to simulate internal recovery attempts.
Outcome: The proposed framework achieves superior trade-off between unlearning efficacy and model utility.
Adversarial Preference Learning for Robust LLM Alignment (2025.findings-acl)

Copied to clipboard

Challenge: Modern language models rely on Reinforcement Learning from Human Feedback (RLHF) to encourage safe behaviors, but they remain vulnerable to adversarial attacks due to three key limitations: (1) the inefficiency and high cost of human annotation; (2) the vast diversity of potential adversarials; and (3) the risk of feedback bias and reward hacking.
Approach: They propose an iterative adversarial training method that incorporates three key innovations to address these challenges.
Outcome: Experiments on Mistral-7B-Instruct-v0.3 show that the proposed method significantly enhances robustness and reduces harmful outputs from 5.88% to 0.43%.
LoPT: Lossless Parallel Tokenization Acceleration for Long Context Inference of Large Language Model (2026.acl-long)

Copied to clipboard

Challenge: Existing parallel tokenization methods suffer from inconsistent results due to boundary artifacts that occur after merging.
Approach: They propose a Lossless Parallel Tokenization framework that ensures output identical to standard sequential tokenization.
Outcome: The proposed method achieves significant speedup while guaranteeing lossless tokenization.
SpeechGPT: Empowering Large Language Models with Intrinsic Cross-Modal Conversational Abilities (2023.findings-emnlp)

Copied to clipboard

Challenge: Existing multi-modal large language models typically adopt the cascade paradigm, preventing inter-modal knowledge transfer.
Approach: They propose a large language model with intrinsic cross-modal conversational abilities . they construct a cross-text speech instruction dataset and employ a three-stage training strategy .
Outcome: The proposed model can follow cross-modal human instructions and handle multiple modalities with one model.
NeuralClassifier: An Open-source Neural Hierarchical Multi-label Text Classification Toolkit (P19-3)

Copied to clipboard

Challenge: NeuralClassifier is a toolkit for hierarchical multi-label text classification.
Approach: They propose a toolkit for neural hierarchical multi-label text classification . they use a variety of text encoders to implement the model .
Outcome: The proposed model achieves comparable performance with reported results in the literature.
HER: Human-like Reasoning and Reinforcement Learning for LLM Role-playing (2026.findings-acl)

Copied to clipboard

Challenge: Existing models for LLM role-playing lack high-quality datasets with explicit reasoning traces and reliable reward signals aligned with human preferences.
Approach: They propose a unified framework for cognitive-level persona simulation that strictly distinguishes characters’ first-person thinking processes from LLMs’ third-person reasoning.
Outcome: The proposed framework outperforms the Qwen3-32B baseline model and achieves a 30.26% and 14.97% performance on the minimax benchmarks.
TokDrift: When LLM Speaks in Subwords but Code Speaks in Grammar (2026.acl-long)

Copied to clipboard

Challenge: Large language models (LLMs) for code rely on subword tokenizers learned from mixed natural language text and programming language code but driven by statistics rather than grammar.
Approach: They propose a framework that applies semantic-preserving rewrite rules to create code variants differing only in tokenization.
Outcome: The proposed framework can create code variants differing only in tokenization . the findings highlight the need for grammar-aware tokenization for future code LLMs.
PerturbScore: Connecting Discrete and Continuous Perturbations in NLP (2023.findings-emnlp)

Copied to clipboard

Challenge: Natural language processing (NLP) applications are growing rapidly due to discrete nature of texts.
Approach: They propose to connect discrete perturbations with continuous perturbations to help understand discrete ones in NLP models.
Outcome: The proposed method surpasses methods used in discrete perturbation measuring and can be generalized to different datasets, perturbation methods.
Impact of Evaluation Methodologies on Code Summarization (2022.acl-long)

Copied to clipboard

Challenge: Existing evaluation methodologies for code summarization tasks do not consider timestamps of code and comments.
Approach: They propose a time-segmented evaluation methodology for code summarization that considers timestamps of code and comments during evaluation.
Outcome: The proposed evaluation methodology compares with other evaluation methodologies that have been widely used.
Improving Disentangled Text Representation Learning with Information-Theoretic Guidance (2020.acl-main)

Copied to clipboard

Challenge: Disentangled representation learning (DRL) maps different aspects of data into distinct and independent low-dimensional latent vector spaces.
Approach: They propose a method that manifests disentangled representations of text without supervision on semantics by minimizing the upper bound between style and content.
Outcome: The proposed method improves on conditional text generation and text-style transfer tasks and improves style preservation.
NewsBench: A Systematic Evaluation Framework for Assessing Editorial Capabilities of Large Language Models in Chinese Journalism (2024.acl-long)

Copied to clipboard

Challenge: a novel evaluation framework assesses the capabilities of Large Language Models (LLMs) for editorial capabilities in Chinese journalism.
Approach: They propose to use a benchmark dataset to assess the capabilities of Large Language Models (LLMs) for editorial capabilities in Chinese journalism.
Outcome: The proposed evaluation framework is based on a dataset of 1,267 test samples in 24 news domains.
UnifiedVisual: A Framework for Constructing Unified Vision-Language Datasets (2025.emnlp-main)

Copied to clipboard

Challenge: Existing datasets address understanding and generation in isolation, limiting the performance of unified vision large language models.
Approach: They propose a dataset that facilitates mutual enhancement between multimodal understanding and generation.
Outcome: The proposed framework integrates diverse visual and textual inputs and outputs, enabling comprehensive cross-modal reasoning and precise text-to-image alignment.
Case2Code: Scalable Synthetic Data for Code Generation (2025.coling-main)

Copied to clipboard

Challenge: Large Language Models (LLMs) have shown outstanding breakthroughs in code generation.
Approach: They propose a case-to-code induction task that exploits the expressiveness and correctness of programs by incorporating LLMs into their training.
Outcome: The proposed task improves distribution case-to-code induction and various coding generation tasks.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations