Papers by Yaxin Xu

7 papers
Legal Mathematical Reasoning with LLMs: Procedural Alignment through Two-Stage Reinforcement Learning (2025.findings-emnlp)

Copied to clipboard

Challenge: Existing legal mathematical reasoning models lack structured numerical reasoning . existing models perform poorly on LexNum, while LexPam improves both mathematical accuracy and legal coherence.
Approach: They propose a legal mathematical reasoning benchmark LexNum and LexPam to address this problem . LexPam is a two-stage reinforcement learning framework for efficient legal reasoning training.
Outcome: The proposed framework improves mathematical accuracy and legal coherence . it also improves legal cohesion and generalizes effectively across tasks and domains.
TCPO: Thought-Centric Preference Optimization for Effective Embodied Decision-making (2025.emnlp-main)

Copied to clipboard

Challenge: Existing post-SFT methods for embodied AI are constrained by sparse rewards and action-only optimization, resulting in low sample efficiency, poor consistency, and model degradation.
Approach: They propose to integrate Thought-Centric Preference Optimization (TCPO) into embodied decision-making by transforming sparse reward signals into richer step sample pairs.
Outcome: The proposed approach achieves an average success rate of 26.67% in the ALFWorld environment, and a 6% improvement over RL4VLM.
Exploring Memorization in Fine-tuned Language Models (2024.acl-long)

Copied to clipboard

Challenge: Existing studies have shown that pre-trained langauge models tend to memorize and regenerate segments of their pre-training corpus when prompted appropriately.
Approach: They conduct the first comprehensive analysis to explore language models’ memorization during fine-tuning across tasks.
Outcome: The proposed analysis shows that memorization presents a strong disparity among different fine-tuning tasks.
“Yes, My LoRD.” Guiding Language Model Extraction with Locality Reinforced Distillation (2025.acl-long)

Copied to clipboard

Challenge: Existing methods for model extraction attacks on large language models are inadequate . existing methods neglect the inconsistency between training tasks and LLM alignment .
Approach: They propose a model extraction algorithm that uses a policy-gradient-style training task to guide the crafting of preference for the local model.
Outcome: The proposed algorithm reduces query complexity while mitigating watermark protection . it can extract various state-of-the-art commercial LLMs while minimizing query complexity .
ChatVLA: Unified Multimodal Understanding and Robot Control with Vision-Language-Action Model (2025.emnlp-main)

Copied to clipboard

Challenge: Recent advances in vision-language-action models prioritize robotic action mastery . however, models trained on visual-text pairs struggle to interpret multimodal data .
Approach: They propose a framework that integrates multimodal data after initial control mastery and a Mixture-of-Experts architecture to minimize task interference.
Outcome: The proposed framework surpasses state-of-the-art vision-language-action (VLA) methods on multimodal understanding benchmarks and achieves six times higher performance on visual question-answering datasets.
Data Augmentation with Adversarial Training for Cross-Lingual NLI (2021.acl-long)

Copied to clipboard

Challenge: Existing approaches to train cross-lingual models with labeled data are subpar, resulting in subpar results.
Approach: They propose a data augmentation strategy that enriches data to reflect more diversity in a semantically faithful way and leverages adversarial training regimens to achieve greater robustness.
Outcome: The proposed approach improves cross-lingual inference by leveraging the data to reflect more diversity in a semantically faithful way.
MCP-Flow: Facilitating LLM Agents to Master Real-World, Diverse and Scaling MCP Tools (2026.acl-long)

Copied to clipboard

Challenge: Existing research on Large Language Models (LLMs) relies on few servers and lacks training support.
Approach: They propose a web-agent-driven pipeline for large-scale server discovery, data synthesis, and model training that collects and filters data from 1166 servers and 11536 tools.
Outcome: Empirical evidence shows that MCP-Flow generates higher quality instruction-function call pairs and higher agentic task performance than previous work.

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