5 papers
Unlocking Token Rewards via Training-Free Reward Attribution
Sitong Wu, Haoru Tan, Bin Xia +6
In this paper, we propose an extremely efficient, training-free method to extract token-level reward signals directly from an existing deep reward model. Our core idea is to attrib…
Dynamic Important Example Mining for Reinforcement Finetuning
Haoru Tan, Sitong Wu, Yanfeng Chen +9
Reinforcement fine-tuning (RFT) is increasingly used to strengthen the reasoning abilities of large models, yet its effectiveness is bound by how training data are selected and use…
Dataset Distillation by Influence Matching
Haoru Tan, Wang Wang, Sitong Wu +5
We revisit dataset distillation from an outcome-centric perspective. Rather than aligning process surrogates (per-step gradients or training trajectories), Influence Matching (Inf-…
Moment-Video: Diagnosing Temporal Fidelity of Video MLLMs on Momentary Visual Events
Xiaolin Liu, Yilun Zhu, Xiangyu Zhao +9
Video multimodal large language models (MLLMs) have made rapid progress in general and long-form video understanding, yet their ability to capture and use brief, answer-critical vi…
Hunyuan-TurboS: Advancing Large Language Models through Mamba-Transformer Synergy and Adaptive Chain-of-Thought
Tencent Hunyuan Team, Ao Liu, Botong Zhou +248
As Large Language Models (LLMs) rapidly advance, we introduce Hunyuan-TurboS, a novel large hybrid Transformer-Mamba Mixture of Experts (MoE) model. It synergistically combines Mam…