Metraipolis › archive › dc_ca7be871aeb2

Robust Failure, Conservative Repair: Textual Knowledge Distillation from Cross-Model Failures — arXiv:2609.25400

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.25400
License (as archived)
CC0-1.0
Archived
2026-10-08T13:44:57.114Z
SHA-256 of the archived text
ac8f91369813d6d5ec5693004a9e1e9c8f0915ddf5a10225619bb0202875741d
Stable URL of this page
https://metraipolis.com/documento/dc_ca7be871aeb2

How to cite

Robust Failure, Conservative Repair: Textual Knowledge Distillation from Cross-Model Failures — arXiv:2609.25400 — arXiv (en), https://arxiv.org/abs/2609.25400 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ca7be871aeb2 (Metraipolis archive dc_ca7be871aeb2, sha256 ac8f91369813d6d5ec5693004a9e1e9c8f0915ddf5a10225619bb0202875741d)
BibTeX
@misc{dcca7be871aeb2,
  title = {Robust Failure, Conservative Repair: Textual Knowledge Distillation from Cross-Model Failures — arXiv:2609.25400},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.25400},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ca7be871aeb2; sha256 ac8f91369813d6d5ec5693004a9e1e9c8f0915ddf5a10225619bb0202875741d},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.25400. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Failure-based textual knowledge distillation aims to discover gaps in a model's knowledge by examining its task errors. The distilled knowledge can be useful for the reasoning of both this model ("source model") and other models. However, this transfer of knowledge may not be stable. We define a rule atom to be a standalone rule injected into a model's textual input at inference time. A rule atom can encode transferable task knowledge or model-specific reasoning patches that can confuse other models. Also, the injected rule atoms can be misapplied to unrelated cases, causing the model to incorrectly flip its answer based on irrelevant information. Building on a pipeline that distills training examples into task-specific cheat sheets that aid model reasoning, we examine when failure-derived rules can improve these cheat sheets. Our early experiment shows rule distillation from a single model's failures underperforms the baseline cheat sheet on non-source model families. This motivates Robust Failure, Conservative Repair (RFCR), a textual distillation procedure that derives rules from failures shared across models, sharpens their application boundaries using boundary cases, and abstains when no useful rule is found. On a 400-item BIG-Bench Hard task set, RFCR improves the baseline cheat sheets from 68.50% to 71.25% (+2.75 pp; 95% CI [+1.25,+4.50]) without performance degradation on previously correct cases. Ablations and cross-model diagnostics support that accuracy gains come from both new knowledge injection and strict rule-application control.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-23 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.25400 — 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 2026-10-08T13:18:38.078Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-23?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).