Papers by Wen Du

21 papers
Learning to Imagine: Visually-Augmented Natural Language Generation (2023.acl-long)

Copied to clipboard

Challenge: Existing methods for natural language generation are pre-trained on text-only corpora, resulting in visual commonsense.
Approach: They propose a method that makes pre-trained language models learn to imagine for visually-augmented natural language generation.
Outcome: The proposed method is compatible with Transformer-based architecture.
Generalization-Enhanced Code Vulnerability Detection via Multi-Task Instruction Fine-Tuning (2024.findings-acl)

Copied to clipboard

Challenge: Existing CodePre-trained models struggle to generalize due to superficial mapping from source code to labels instead of understanding the root causes of code vulnerabilities.
Approach: They propose a framework that integrates multi-task learning with Large Language Models to effectively mine deep-seated vulnerability features.
Outcome: The proposed framework surpasses seven state-of-the-art models in effectiveness, generalization, and robustness.
Evaluating Object Hallucination in Large Vision-Language Models (2023.emnlp-main)

Copied to clipboard

Challenge: Large vision-language models (LVLMs) suffer from object hallucinations, i.e., they tend to generate objects inconsistent with the target images in the descriptions.
Approach: They propose to integrate powerful large vision-language models (LVLMs) they propose a polling-based query method to evaluate object hallucination .
Outcome: The proposed model can evaluate object hallucination in a more stable and flexible way.
QiMeng-PRepair: Precise Code Repair via Edit-Aware Reward Optimization (2026.acl-long)

Copied to clipboard

Challenge: Existing approaches to program repair are based on correctness alone.
Approach: They propose a framework that mitigates over-editing and improves repair accuracy by generating buggy programs and re-edits.
Outcome: The proposed framework improves repair precision by 31.4% under fix1@1, a metric that considers repair correctness and extent, and significantly increases decoding throughput when combined with speculative editing.
Light-R1: Curriculum SFT, DPO and RL for Long COT from Scratch and Beyond (2025.acl-industry)

Copied to clipboard

Challenge: Experimental results show that opensource curriculum training is more effective when distinct datasets are available for different training stages.
Approach: They propose an opensource suite for training long reasoning models using publicdata and models.
Outcome: The proposed model outperforms DeepSeek-R1-DistillQwen-32B models in math reasoning.
DoTAT: A Domain-oriented Text Annotation Tool (2022.acl-demo)

Copied to clipboard

Challenge: DoTAT is a domain-oriented text annotation tool that can reduce the time for event annotation by 19.7% . the tool supports multi-person collaborative process with automatically merging and review .
Approach: They propose a domain-oriented text annotation tool called DoTAT . it provides multi-person collaborative process with automatic merging and review .
Outcome: The proposed tool can reduce the time for event annotation by 19.7% compared with existing tools.
FaithLM: Towards Faithful Explanations for Large Language Models (2026.eacl-long)

Copied to clipboard

Challenge: Large language models (LLMs) produce natural language explanations, but they lack faithfulness and do not reflect the evidence the model uses to decide.
Approach: They propose a model-agnostic framework that evaluates and improves the faithfulness of LLM explanations without token masking or task-specific heuristics.
Outcome: The proposed framework improves faithfulness of large language models without masking or heuristics.
Time-MQA: Time Series Multi-Task Question Answering with Context Enhancement (2025.acl-long)

Copied to clipboard

Challenge: Existing time series models focus on a narrow spectrum of tasks, such as forecasting or anomaly detection.
Approach: They propose a framework that enables natural language queries across multiple time series tasks such as numerical analytical tasks and open-ended question answering with reasoning.
Outcome: The proposed framework enables natural language queries across multiple time series tasks and allows for more advanced and intuitive interactions with temporal data.
OmniFlatten: An End-to-end GPT Model for Seamless Voice Conversation (2025.acl-long)

Copied to clipboard

Challenge: Full-duplex spoken dialogue systems allow simultaneous bidirectional communication . low latency and natural interactions in full-duplice systems remains a challenge .
Approach: They propose a multi-stage post-training scheme that adapts a text large language model into a speech-text dialogue LLM.
Outcome: The proposed model can model human conversation behaviors with low latency and natural interactions with low delay.
Zero-shot Visual Question Answering with Language Model Feedback (2023.findings-acl)

Copied to clipboard

Challenge: Existing methods for knowledge-based visual question answering are based on pre-trained language models.
Approach: They propose a language model guided captioning approach that leverages a pre-trained language model to generate captions for an image to help answer a visual question.
Outcome: The proposed method outperforms several competing methods on the knowledge-based VQA task and achieves comparable results to a fine-tuned VLP model.
Efficient Hyperparameter Optimization for LLM Reinforcement Learning (2026.acl-long)

Copied to clipboard

Challenge: Existing hyperparameter optimization methods are inefficient in reinforcement learning due to model scale and resource-intensive training cycles.
Approach: They propose a hyperparameter optimization method that adapts both model size and training budget as fidelity.
Outcome: The proposed method significantly improves the computational efficiency of each trial (up to 14.9) over existing HPO methods.
Token-Level Policy Optimization: Linking Group-Level Rewards to Token-Level Aggregation via sequence-level likelihood (2026.acl-long)

Copied to clipboard

Challenge: Group Relative Policy Optimization (GRPO) has significantly advanced the reasoning ability of large language models (LLMs).
Approach: They propose a token-level framework that leverages sequence-level likelihood to link group-level rewards with individual tokens via token- level aggregation and introduces a KL-Divergence mask constraint that targets tokens with positive advantages and decreasing entropy to mitigate abrupt policy updates.
Outcome: Experiments show that TEPO achieves state-of-the-art performance on mathematical reasoning benchmarks and reduces convergence time by 50% compared with GRPO/DAPO.
Knowing-how & Knowing-that: A New Task for Machine Comprehension of User Manuals (2023.findings-acl)

Copied to clipboard

Challenge: Existing methods for machine reading comprehension of user manuals have trouble answering complex questions.
Approach: They propose a knowing-how & knowing-that task that requires the model to answer factoid-style, procedure-style and inconsistent questions about user manuals.
Outcome: The proposed model can answer factoid-style, procedure-style and inconsistent questions about user manuals.
f-Divergence Minimization for Sequence-Level Knowledge Distillation (2023.acl-long)

Copied to clipboard

Challenge: Existing knowledge distillation approaches focus on minimizing a generalized f-divergence function.
Approach: They propose a framework which formulates sequence-level knowledge distillation as minimizing a generalized f-divergence function.
Outcome: The proposed framework outperforms existing methods and reduces intractable divergence to word-level losses.
Retrieval-Augmented Process Reward Model for Generalizable Mathematical Reasoning (2025.findings-acl)

Copied to clipboard

Challenge: Large language models (LLMs) have advanced mathematical reasoning, but they still struggle with out-of-distribution (OOD) issues.
Approach: They propose a framework to evaluate the logical validity of reasoning steps . they retrieves semantically similar questions and steps for PRM as a warmup .
Outcome: The proposed framework outperforms baseline models on multiple real-world datasets.
CARE: A Clue-guided Assistant for CSRs to Read User Manuals (2024.acl-long)

Copied to clipboard

Challenge: Current solutions don't fit the online custom service scenarios well due to the lack of attention to user questions and possible responses.
Approach: They propose to build a clue-guided assistant for customer service representations (CSRs) that can provide accurate responses and explicitly show explainable paths about how to arrive at these responses.
Outcome: The proposed assistant can reduce CSRs' reading burden and keep high service quality, in particular with >35% decrease in time spent and keeping a >0.75 ICC score.
Quantification of Large Language Model Distillation (2025.acl-long)

Copied to clipboard

Challenge: Existing studies have revealed the robustness degra-dation caused by data distillation.
Approach: They propose a framework to evaluate and quantify model distillation . they aim to identify identity cognition contradictions and analyse multi-granularity response similarities across models to measure the extent of homogenization.
Outcome: The proposed framework addresses two key aspects: (1) Identifying identity cognition contradictions to assess discrepancies in how models perceive and represent identity-related information; (2) Analyzing multi-granularity response similarities across models to measure the extent of homogenization.
What Makes for Good Visual Instructions? Synthesizing Complex Visual Reasoning Instructions for Visual Instruction Tuning (2025.coling-main)

Copied to clipboard

Challenge: Experimental results show that visual instruction tuning improves performance of Multi-modal Large Language Models (MLLMs) to extend the application scope of Large Language Modells, a surge of work augments LLMs with vision encoders to endow the ability of multi-modal cognition and reasoning.
Approach: They propose a systematic approach to create high-quality visual reasoning instructions using a synthesize-complicate-reformulate paradigm.
Outcome: The proposed method improves performance of MLLMs by 27.86% and 27.60% on MME-Perception and MME Cognition.
Safety-Utility Conflicts Are Not Global: Surgical Alignment via Head-Level Diagnosis (2026.acl-long)

Copied to clipboard

Challenge: Existing mitigation strategies rely on global gradient geometry to resolve alignment conflicts . however, they overlook Modular Heterogeneity within Transformers, resulting in suboptimal trade-offs . Conflict-Aware Sparse Tuning (CAST) combines head-level diagnosis with sparse fine-tuning .
Approach: They propose a framework that integrates head-level diagnosis with sparse fine-tuning to address this limitation.
Outcome: The proposed framework integrates head-level diagnosis with sparse fine-tuning to reduce alignment conflicts in LLMs.
LogToP: Logic Tree-of-Program with Table Instruction-tuned LLMs for Controlled Logical Table-to-Text Generation (2026.findings-eacl)

Copied to clipboard

Challenge: Existing LLMs are difficult to achieve satisfactory results in table-related tasks.
Approach: They propose to develop a specialized logical table-to-text generation model that can be used for table-related tasks.
Outcome: The proposed model achieves state-of-the-art on a Logic2Text dataset.

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