Episode
Distilling Long-CoT Reasoning through Collaborative Step-wise Multi-Teacher Decoding
- Podcast
- Daily Paper Cast
- Published
- May 19, 2026
- Duration seconds
- 1272
- Processing state
not_requested- Canonical source
- https://share.transistor.fm/s/e1a410e1
Actions
POST https://stenobird.com/v1/public/podcasts/daily-paper-cast-7079649/episodes/distilling-long-cot-reasoning-through-collaborative-step-wise-multi-teacher-decoding/transcription-requests
Idempotently request low-priority transcript generation for this episode.GET https://stenobird.com/podcast/daily-paper-cast-7079649/distilling-long-cot-reasoning-through-collaborative-step-wise-multi-teacher-decoding.md
Read the agent-friendly Markdown representation of this episode resource.
Summary
đŸ¤— Upvotes: 34 | cs.AI Authors: Taewon Yun, Jisu Shin, Jeonghwan Choi, Seunghwan Bang, Hwanjun Song Title: Distilling Long-CoT Reasoning through Collaborative Step-wise Multi-Teacher Decoding Arxiv: http://arxiv.org/abs/2605.02290v1 Abstract: Distilling large reasoning models is essential for making Long-CoT reasoning practical, as full-scale inference remains computationally prohibitive. Existing curation-based approaches select complete reasoning traces post-hoc, overlooking collaboration among heterogeneous teachers and lacking dynamic exploration, which leads to redundant sampling and missed complementary reasoning. We introduce CoRD, a collaborative multi-teacher decoding framework that performs step-wise reasoning synthesis guided by predictive perplexity-based scoring and beam search. This enables heterogeneous LRMs to jointly construct coherent reasoning trajectories while efficiently preserving diverse, high-potential hypotheses. Experiments show that CoRD produces higher-quality reasoning data and achieves near teacher-level student performance with fewer, structured supervision signals, without substantial efficiency overhead. CoRD further generalizes well to out-of-domain and open-ended settings. The dataset and model are available at \href{https://github.com/DISL-Lab/CoRD}{https://github.com/DISL-Lab/CoRD}.