Metraipolis › archive › dc_d9f5bf5cd367

SPECTRUM: Proximal Spectral Modulation for Looped Self-Distillation — arXiv:2610.07237

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.07237
License (as archived): CC0-1.0
Archived: 2026-10-07T11:18:27.017Z
SHA-256 of the archived text: c5f6c594b2f12a8056b9c004e4f0172e2e9cdf14c5993d13a5bccbc6a3e67a63
Stable URL of this page: https://metraipolis.com/documento/dc_d9f5bf5cd367

How to cite

SPECTRUM: Proximal Spectral Modulation for Looped Self-Distillation — arXiv:2610.07237 — arXiv (en), https://arxiv.org/abs/2610.07237 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d9f5bf5cd367 (Metraipolis archive dc_d9f5bf5cd367, sha256 c5f6c594b2f12a8056b9c004e4f0172e2e9cdf14c5993d13a5bccbc6a3e67a63)

BibTeX

@misc{dcd9f5bf5cd367,
  title = {SPECTRUM: Proximal Spectral Modulation for Looped Self-Distillation — arXiv:2610.07237},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.07237},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d9f5bf5cd367; sha256 c5f6c594b2f12a8056b9c004e4f0172e2e9cdf14c5993d13a5bccbc6a3e67a63},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.07237. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A model that learns from its own outputs inherits more than their correctness: it inherits which solutions it produces. We formulate Looped Self-Distillation, a self-evolution framework for code generation in which a model repeatedly generates and learns from its own raw outputs, under a fixed information budget, without ongoing external assessment or test-based selection of the generated samples. We identify a consequential separation: correctness can improve while the breadth of correct implementations contracts. We introduce SPECTRUM, which re-estimates loss-sensitive key/value geometry from a fixed reference anchor at each round and converts it into full-rank proximal spectral modulation. All generated completions train a single student, whose subsequent inference requires no intervention. After five rounds of experiments on MBPP, SPECTRUM retains 89.9% of the initial model's 64-sample correct AST richness, compared with 66.4% for Vanilla self-distillation and 65.5% for a subspace-projection control. The advantage persists at matched correct-sample counts. Without further training or recalibration, the resulting student also achieves higher matched-correct richness than Vanilla SD on HumanEval+ and APPS Intro, demonstrating transfer of the diversity benefit. These findings establish correct-solution retention as a complementary objective of recursive self-improvement (RSI) and show that generation-time intervention can improve the solution repertoire retained by subsequent students.

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/2610.07237 — 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 2026-10-07T11:15:24.486Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).