arXiv AI July 21, 2026 By Satyam Kumar, Saurabh Jha

CADENCE: Closing the Reasoning Gap via Coverage-Adaptive On-Policy Distillation

arXiv:2607. 16955v1 Announce Type: cross Abstract: On-policy knowledge distillation transfers reasoning from large teachers to compact students, but existing approaches suffer three compounding failure modes: (i) cold-start collapse, where a fresh student assigns near-zero mass to teacher-preferred tokens; (ii) state-agnostic divergence scheduling, where time-only forward/reverse-KL interpolation ignores the student's coverage state; and (iii) binary reward sparsity, where pass/fail signals discard information from partially correct traces.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

efficiency

arXiv Computation and Language

4d ago

Learning from Teacher Continuations at Student States

arXiv:2609.36246v1 Announce Type: new Abstract: We present OLIVE (OnLine InterVEntion). At each iteration, the evolving student policy generates a new prefix, the teacher continues it autoregressivel...

By Haojin Wang, Dylan Zhang, Huaibo Chen, Suhao Yu, Yihang Sun, Zhanyang Jin, Jiaying Ye, Dianqi Li, Prasanna Sattigeri, Kamal Youcef-Toumi, Hao Peng

llms agents fine-tuning efficiency

arXiv AI

3d ago

STEPS: Selective On-Policy Self-Distillation for Reasoning

arXiv:2605.10194v2 Announce Type: replace Abstract: On-policy self-distillation (OPSD) uses a model as its own teacher under privileged context, providing token-level supervision on the model's own r...

By Jiaxuan Wang, Xuan Ouyang, Zhiyu Chen, Yulan Hu, Lan-Zhe Guo

reinforcement-learning efficiency benchmarks

arXiv Computation and Language

4d ago

Distilling What Matters: Confidence-Aware Selective Distillation for Large Language Models

arXiv:2609.36734v1 Announce Type: new Abstract: Knowledge Distillation (KD) trains a smaller-capacity student model to imitate a larger-capacity teacher model by matching output distributions, implic...

By Ayan Sengupta, Vaibhav Seth, Tanmoy Chakraborty

llms efficiency benchmarks safety

arXiv Computer Vision

3d ago

ReDiF: Resource-Efficient Few-Step Diffusion Distillation via Reinforcement Learning

arXiv:2512.22802v2 Announce Type: replace-cross Abstract: Step distillation accelerates diffusion sampling by training a few-step student to imitate a many-step teacher, but distillation itself remai...

By Amirhossein Tighkhorshid, Zahra Dehghanian, Hamid R. Rabiee

diffusion reinforcement-learning efficiency safety

arXiv AI

Sep 17

Trajectory Learnability for Offline On-Policy Distillation with Imperfect Teachers

The paper introduces a method for offline on‑policy distillation that addresses the problem of imperfect teacher supervision. By training on teacher‑successful problems and measuring changes in token likelihoods on teacher‑failed trajectories, the authors derive a learnability signal that weights the distillation loss. This approach improves performance on mathematical reasoning and code generation tasks while reducing computational cost compared to online distillation.

By Yihao Ai, Weilong Yan

efficiency benchmarks

arXiv AI

4d ago

Guide, Then Let Go: Gap-Adaptive Teacher Scheduling for Sparse-Reward Agentic RL

arXiv:2609.37898v1 Announce Type: new Abstract: Reinforcement learning for long-horizon agents typically relies on sparse outcome-based rewards. This leads to a severe cold-start problem, as early-st...

By Youling Huang, Tiankuo Xu, Jiaji Liu, Tong Zheng, Shuo Zhou, Shaotong Qi, Junchi Yao, Shiyang Liu, Hao Xu, Pengcheng Xu, Bo Huang, Hongyi Fu, Lin Lin

agents reinforcement-learning efficiency

One email a morning, machine-written

Related stories

Learning from Teacher Continuations at Student States

STEPS: Selective On-Policy Self-Distillation for Reasoning

Distilling What Matters: Confidence-Aware Selective Distillation for Large Language Models

ReDiF: Resource-Efficient Few-Step Diffusion Distillation via Reinforcement Learning

Trajectory Learnability for Offline On-Policy Distillation with Imperfect Teachers

Guide, Then Let Go: Gap-Adaptive Teacher Scheduling for Sparse-Reward Agentic RL