Metraipolis › archive › dc_4301de4633d7
Coupled Calibration and Learning: Mitigating Teacher Bias in LLM Distillation without Target-Domain Reward Feedback — arXiv:2609.17474
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.17474
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-15T18:38:59.284Z
- SHA-256 of the archived text
5cf99d2958971d6a6f20b58eb0327623a026d4f57e9e9a4e6f72434aef392e5f- Stable URL of this page
https://metraipolis.com/documento/dc_4301de4633d7
How to cite
Coupled Calibration and Learning: Mitigating Teacher Bias in LLM Distillation without Target-Domain Reward Feedback — arXiv:2609.17474 — arXiv (en), https://arxiv.org/abs/2609.17474 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4301de4633d7 (Metraipolis archive dc_4301de4633d7, sha256 5cf99d2958971d6a6f20b58eb0327623a026d4f57e9e9a4e6f72434aef392e5f)
BibTeX
@misc{dc4301de4633d7,
title = {Coupled Calibration and Learning: Mitigating Teacher Bias in LLM Distillation without Target-Domain Reward Feedback — arXiv:2609.17474},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.17474},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4301de4633d7; sha256 5cf99d2958971d6a6f20b58eb0327623a026d4f57e9e9a4e6f72434aef392e5f},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.17474. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Large language model (LLM) distillation aims to transfer the capabilities of a powerful teacher to a smaller student. Direct imitation, however, can also transfer the teacher's systematic bias and errors. This challenge is particularly pronounced under covariate shift, when the teacher's reliability on target questions is uncertain and target-domain reward feedback is unavailable. We propose Coupled Calibration and Learning (CCL), an LLM distillation algorithm that couples teacher calibration with student updates through token-level branching, using reward feedback only on source questions. Each iteration calibrates the teacher using source feedback and then uses the calibrated teacher to train the student on target questions. The updated student, in turn, informs subsequent calibration. In an autoregressive policy framework, we prove that the output student's expected average Kullback-Leibler divergence to the oracle student converges to zero at a polynomial rate in the number of iterations. The oracle maximizes the true reference-regularized target reward within the student class, which need not represent the unrestricted optimal policy. Our analysis quantifies the progress of projected student gradient updates while controlling the error in teacher calibration. We further establish a separation from regularized direct matching: its error relative to the oracle student can remain bounded away from zero even when the teacher achieves higher regularized target reward than every student policy. These results demonstrate that LLM distillation can overcome persistent teacher bias and recover the optimal student through coupled calibration and learning, without target-domain reward feedback. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-16 in cs.AI). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Statistics Theory (math.ST); Machine Learning (stat.ML). Pagina del lavoro: https://arxiv.org/abs/2609.17474 — 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 2026-10-15T18:03:09.760Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-16?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).