6 papers
Efficient Multilingual Reasoning Transfer via Progressive Code-Switching
Zhijun Wang, Junxiao Liu, Hao Zhou +3
Large reasoning models (LRMs) have achieved strong reasoning capabilities in English, yet their performance degrades significantly when required to reason in other languages. A nat…
NaviAgent: Graph-Driven Bilevel Planning for Scalable Tool Orchestration
Yan Jiang, Hao Zhou, Lizhong GU +4
Large Language Models (LLMs) increasingly act as function-call agents that invoke external tools to tackle tasks beyond their static knowledge. However, they typically invoke tools…
Investigating Cross-Modal Skill Injection: Scenarios, Methods, and Hyperparameters
Zhiyu Xu, Lean Wang, Yuanxin Liu +5
Vision-Language Models (VLMs) have demonstrated remarkable proficiency in general multi-modal understanding; yet they struggle to efficiently acquire continually evolving domain-sp…
A Data-Efficient Path to Multilingual LLMs: Language Expansion via Post-training PARAM Integration into Upcycled MoE
Hao Zhou, Tianhao Li, Zhijun Wang +6
Expanding Large Language Models~(LLMs) to new languages is a costly endeavor, demanding extensive Continued Pre-Training~(CPT) and data-intensive alignment. While recent data-free…
Fidelity Probes for Specification--Code Alignment
Ferhat Erata, Hao Zhou, Luke Huan
We introduce fidelity probes: natural-language questions generated from a reference artifact with code-derived ground-truth answers, answered from a candidate specification. The fr…
Retrieval is Not Enough: Enhancing RAG Reasoning through Test-Time Critique and Optimization
Jiaqi Wei, Hao Zhou, Xiang Zhang +6
Retrieval-augmented generation (RAG) has become a widely adopted paradigm for enabling knowledge-grounded large language models (LLMs). However, standard RAG pipelines often fail t…