Metraipolis › archive › dc_ef4ff79adfe5

ReTeach: Building a Self-Teacher through Multi-Round Reflection and Retry — arXiv:2610.11529

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.11529
License (as archived): CC0-1.0
Archived: 2026-10-09T05:48:00.609Z
SHA-256 of the archived text: 62b9a8a59b8df75c19491892d215562610fb5cf2a9d2bcfac339d49d2ef05ec3
Stable URL of this page: https://metraipolis.com/documento/dc_ef4ff79adfe5

How to cite

ReTeach: Building a Self-Teacher through Multi-Round Reflection and Retry — arXiv:2610.11529 — arXiv (en), https://arxiv.org/abs/2610.11529 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ef4ff79adfe5 (Metraipolis archive dc_ef4ff79adfe5, sha256 62b9a8a59b8df75c19491892d215562610fb5cf2a9d2bcfac339d49d2ef05ec3)

BibTeX

@misc{dcef4ff79adfe5,
  title = {ReTeach: Building a Self-Teacher through Multi-Round Reflection and Retry — arXiv:2610.11529},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.11529},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ef4ff79adfe5; sha256 62b9a8a59b8df75c19491892d215562610fb5cf2a9d2bcfac339d49d2ef05ec3},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.11529. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Self-distillation can improve reasoning without a separately trained, more capable teacher, but its effectiveness depends on how the self-teacher gains an advantage over the student. Conditioning the teacher on reference answers or solutions can provide such an advantage, but this information may be unavailable. Reflection offers a way to derive explicit error diagnoses and revision guidance from self-generated attempts, yet existing reflection-based methods often combine it with reference information, rich task feedback, or persistent memory. We introduce ReTeach, a Reflective self-distillation framework that constructs its self-Teacher through multi-round reflection and retry using only self-generated attempts and outcome-level verification. Starting from an unsuccessful student rollout, the teacher alternates explicit reflection with renewed attempts until success or the retry budget is exhausted, without reference answers or solutions, external diagnostic feedback, or cross-example memory. Each failed retry informs subsequent reflection, while successful correction provides outcome-level evidence for the potential utility of the resulting teacher context. An outcome-aware selection and weighting strategy distinguishes initially correct, reflection-corrected, and unresolved examples, assigning separate weights to their category-normalized distillation losses. Through on-policy distillation, the student matches the teacher's context-conditioned token-level predictive distributions at prefixes of its own rollouts, transferring the benefits of iterative correction while retaining single-pass inference. Across six benchmarks spanning mathematical reasoning, science question answering, and tool use, ReTeach improves average accuracy over GRPO by 1.39 percentage points.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.11529 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-09T05:47:00.799Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).