10 papers
Rep2Skill: Representation-Guided Skill Self-Evolution for LLM Agents
Kaixing Zhang, Changming Li, Yingdong Shi +5
Textual skills enable large language model (LLM) based agents to accumulate reusable procedural knowledge without updating model parameters. Yet existing skill evolution remains la…
OPMD: On-Policy Mechanism Distillation Beyond Output Supervision
Yingdong Shi, Zhihong Li, Changming Li +3
On-policy distillation (OPD) improves language-model by supervising a student on prefixes sampled from its own policy, but its supervision remains confined to the output space. Con…
From Judgment Quality to Downstream Utility: Rethinking LLM-as-a-Judge for Open-Ended Tasks
Zheng Zhang, Lufei Li, Xinyue Tan +4
LLM-as-a-Judge is increasingly used to evaluate policy responses on open-ended tasks that lack ground-truth answers. Existing work often directly converts the resulting judgments i…
Train Ahead, Distill Back: Bootstrapping On-Policy Self-Distillation for Large Language Models
Zheng Zhang, Xinyue Tan, Lufei Li +3
On-policy self-distillation (OPSD) improves large language models by letting a self-teacher with privileged information provide dense token-level supervision on the model's own tra…
AVA-Encoder: Towards Agent-Native Video Representation Learning
Chuyue Li, Jinpeng Yu, Haozhe Wang +7
Agentic video creation systems remain far from cinematic-grade quality: their underlying models struggle to plan how story, characters, visuals, and sound fit together, and high-qu…
UniSteer: Text-Guided Flow Matching in Activation Space for Versatile LLM Steering
Yingdong Shi, Ruiming Zhang, Changming Li +4
Activation-based control steers large language models (LLMs) by intervening on their internal representations during inference, and has emerged as an effective paradigm for control…
KairosAgent: Agentic Time Series Forecasting with Fused Semantic Reasoning
Kun Feng, Ziwei Shan, Yuchen Fang +6
Cross-domain multimodal time series forecasting is a challenging task, requiring models to integrate precise numerical comprehension, cross-domain semantic understanding, and effec…
Large Language Models Explore by Latent Distilling
Yuanhao Zeng, Ao Lu, Lufei Li +3
Generating diverse responses is crucial for test-time scaling of large language models (LLMs), yet standard stochastic sampling mostly yields surface-level lexical variation, limit…
Interpreting and Controlling LLM Reasoning through Integrated Policy Gradient
Changming Li, Kaixing Zhang, Haoyun Xu +4
Large language models (LLMs) demonstrate strong reasoning abilities in solving complex real-world problems. Yet, the internal mechanisms driving these complex reasoning behaviors r…
Grad2Reward: From Sparse Judgment to Dense Rewards for Improving Open-Ended LLM Reasoning
Zheng Zhang, Ao Lu, Yuanhao Zeng +5
Reinforcement Learning with Verifiable Rewards (RLVR) has catalyzed significant breakthroughs in complex LLM reasoning within verifiable domains, such as mathematics and programmin…