Followed topics

Search

Showing top 46 results for "AI agents rollout"

All sources huggingface.co 46

Paper page - Rebellious Student: Reversing Teacher Signals for Reasoning Exploration with Self-Distilled RLVR

…While this guidance is useful when the student has failed, on successful rollouts, the same mechanism instead overwrites the student's choices and suppresses it's own reasoning. Therefore, we propose reading…

Paper page - Memory-Bound but Not Bandwidth-Limited: The Physical AI Inference Gap in Batch-1 LLM Decode

…Generated by Qwen/Qwen2.5-Coder-32B-Instruct Physical AI systems, including robots, autonomous vehicles, embodied agents and edge copilots, often run a different inference workload from cloud LLM serving: single-stream…

Paper page - Beyond GRPO and On-Policy Distillation: An Empirical Sparse-to-Dense Reward Principle for Language-Model Post-Training

…a forward-KL warmup on teacher rollouts followed by OPD on student rollouts is consistently strongest on MATH before any post-bridge student-side sparse RL , and also gives the best pre…

Paper page - LEAD: Length-Efficient Adaptive and Dynamic Reasoning for Large Language Models

…Furthermore, it estimates an adaptive per-problem target length online based on the model's own correct rollouts, applying a symmetric efficiency reward that penalizes both overthinking and over-compression. Evaluated on…

Paper page - Crosslingual On-Policy Self-Distillation for Multilingual Reasoning

…Training minimizes full-distribution token-level divergence on the student's own rollouts, providing dense supervision while avoiding the sparsity and instability of outcome-only reinforcement learning (RL). Experiments on 17 low…

Paper page - Rubric-based On-policy Distillation

…Specifically, ROPD induces prompt-specific rubrics from teacher-student contrasts, and then utilizes these rubrics to score the student rollouts for on-policy optimization. Empirically, ROPD outperforms the advanced logit-based OPD…

Paper page - One Turn Too Late: Response-Aware Defense Against Hidden Malicious Intent in Multi-Turn Dialogue

…To further support training and evaluation, we construct the Multi-Turn Intent Dataset (MTID), which contains branching attack rollouts, matched benign hard negatives , and annotations of the earliest harm-enabling turns. We…

Paper page - Adaptive Teacher Exposure for Self-Distillation in LLM Reasoning

…AI-generated summary On-policy self-distillation has become a strong recipe for LLM reasoning, where a privileged teacher supervises the student's own rollouts while conditioning on the reference solution. A…

Paper page - CausalCine: Real-Time Autoregressive Generation for Multi-Shot Video Narratives

…Trained primarily for short-horizon continuation, they treat long sequences as extended single shots, inevitably suffering from motion stagnation and semantic drift during long rollouts. To bridge this gap, we introduce CausalCine…

Paper page - Beyond Reasoning: Reinforcement Learning Unlocks Parametric Knowledge in LLMs

…those whose answers never appear in 128 pre-RL samples (only ~18% of training data) drive ~83% of the gain, since rare correct rollouts still emerge during training and get reinforced. Together…