Metraipolis › archive › dc_1d2399208ca0

Decoupled Marginal Sharpening for Training-Free Inference-Time Scaling — arXiv:2610.20483

Source

arXiv (en)

Original

https://arxiv.org/abs/2610.20483

License (as archived)

CC0-1.0

Archived

2026-10-15T14:20:35.778Z

SHA-256 of the archived text

a878b930586273d9ddc088089555079e224745c10640b1ab2a532761bb38be74

Stable URL of this page

https://metraipolis.com/documento/dc_1d2399208ca0

How to cite

Decoupled Marginal Sharpening for Training-Free Inference-Time Scaling — arXiv:2610.20483 — arXiv (en), https://arxiv.org/abs/2610.20483 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_1d2399208ca0 (Metraipolis archive dc_1d2399208ca0, sha256 a878b930586273d9ddc088089555079e224745c10640b1ab2a532761bb38be74)

@misc{dc1d2399208ca0,
  title = {Decoupled Marginal Sharpening for Training-Free Inference-Time Scaling — arXiv:2610.20483},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.20483},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_1d2399208ca0; sha256 a878b930586273d9ddc088089555079e224745c10640b1ab2a532761bb38be74},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.20483. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We study how to improve LLM reasoning at inference time without updating model weights or using external reward models. Recent reward-free power-sampling methods define the sequence-level power target $\pi_\alpha(y\mid x)\propto p_0(y\mid x)^\alpha$, where $\alpha>1$, which increases the relative probability of completions that are already likely under the base model. Existing methods use Metropolis--Hastings or Sequential Monte Carlo (SMC) to approximately sample from this target. However, two challenges arise at different levels. At the target level, full-sequence sharpening can concentrate probability on a few exact reasoning--answer trajectories, suppressing alternative paths that support the same answer. At the sampler level, finite-particle SMC can suffer from path collapse when resampling repeatedly duplicates a few high-weight prefixes. To address these challenges, we propose Decoupled Marginal Sharpening (DMS), which separately controls how strongly the model favors particular reasoning paths and how strongly it favors final answers after aggregating support across those paths. At the target level, DMS first samples from a sharpened reasoning distribution, then marginalizes the model's answer probabilities over the resulting weighted reasoning population and sharpens the induced answer marginal. Separate exponents allow mild reasoning sharpening to encourage exploration of diverse paths, followed by stronger answer-marginal sharpening to concentrate probability on answers with greater aggregate support across those paths. At the sampler level, we develop Fully Adapted Annealed SMC, which combines an annealed sequence of power targets with locally fully adapted proposals and predictive ancestor resampling. Experiments on MATH500, GPQA-Diamond, HumanEval, and AIME24/25 across three base models show that DMS achieves the highest mean pass@1 among the compared methods in all tasks.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.20483 — 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 2026-10-15T13:53:35.760Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).