Yixia Li

VFA: Empowering Multilingual MLLMs via Vision-Free Adaptation

Yixia Li, Yaqing Shi, Zhiwen Ruan +6

Multimodal large language models have advanced rapidly, yet most remain English-centric, as scaling multilingual multimodal instruction tuning is limited by the scarcity and high c…

Bridging the Agent-World Gap: Text World Models for LLM-based Agents

Yixia Li, Hongru Wang, Peng Lai +13

Large language model (LLM)-based agents are increasingly used in interactive textual environments, from web navigation and code editing to tool use and long-horizon dialogue. Yet m…

PatchWorld: Gradient-Free Optimization of Executable World Models for Agent Environments

Jiaxin Bai, Yue Guo, Yifei Dong +13

World models for interactive text agents must typically be learned from observation-action trajectories alone. Specifically, the environment returns text observations after each ac…

SPPO: Sequence-Level PPO for Long-Horizon Reasoning Tasks

Tianyi Wang, Yixia Li, Long Li +6

Proximal Policy Optimization (PPO) is central to aligning Large Language Models (LLMs) in reasoning tasks with verifiable rewards. However, standard token-level PPO struggles in th…

Rethinking the Role of Entropy in Optimizing Tool-Use Behaviors for Large Language Model Agents

Zeping Li, Hongru Wang, Yiwen Zhao +7

Tool-using agents based on Large Language Models (LLMs) excel in tasks such as mathematical reasoning and multi-hop question answering. However, in long trajectories, agents often…

Towards Fair and Comprehensive Evaluation of Routers in Collaborative LLM Systems

Wanxing Wu, He Zhu, Yixia Li +7

Large language models (LLMs) have achieved success, but cost and privacy constraints necessitate deploying smaller models locally while offloading complex queries to cloud-based mo…

Anchored Policy Optimization: Mitigating Exploration Collapse Via Support-Constrained Rectification

Tianyi Wang, Long Li, Hongcan Guo +5

Reinforcement Learning with Verifiable Rewards (RLVR) is increasingly viewed as a tree pruning mechanism. However, we identify a systemic pathology termed Recursive Space Contracti…

From Abstract to Contextual: What LLMs Still Cannot Do in Mathematics

Bowen Cao, Dongdong Zhang, Yixia Li +8

Large language models now solve many benchmark math problems at near-expert levels, yet this progress has not fully translated into reliable performance in real-world applications.…

No More Stale Feedback: Co-Evolving Critics for Open-World Agent Learning

Zhicong Li, Lingjie Jiang, Yulan Hu +7

Critique-guided reinforcement learning (RL) has emerged as a powerful paradigm for training LLM agents by augmenting sparse outcome rewards with natural-language feedback. However,…

From Word to World: Can Large Language Models be Implicit Text-based World Models?

Yixia Li, Hongru Wang, Jiahao Qiu +7

Agentic reinforcement learning increasingly relies on experience-driven scaling, yet real-world environments remain non-adaptive, limited in coverage, and difficult to scale. World…

G2: Guided Generation for Enhanced Output Diversity in LLMs

Zhiwen Ruan, Yixia Li, Yefeng Liu +5

Large Language Models (LLMs) have demonstrated exceptional performance across diverse natural language processing tasks. However, these models exhibit a critical limitation in outp…

Enhancing Large Language Model Reasoning via Selective Critical Token Fine-Tuning

Zhiwen Ruan, Yixia Li, He Zhu +4

Large language models (LLMs) primarily rely on supervised fine-tuning (SFT) as a key method to adapt pre-trained models to domain-specific tasks such as mathematical reasoning. How…

VisCodex: Unified Multimodal Code Generation via Merging Vision and Coding Models

Lingjie Jiang, Shaohan Huang, Xun Wu +3

Multimodal large language models (MLLMs) have significantly advanced the integration of visual and textual understanding. However, their ability to generate code from multimodal in…

ImPart: Importance-Aware Delta-Sparsification for Improved Model Compression and Merging in LLMs

Yan Yang, Yixia Li, Hongru Wang +4

With the proliferation of task-specific large language models, delta compression has emerged as a method to mitigate the resource challenges of deploying numerous such models by ef…