There is a newer version of the record available.

Published June 20, 2026 | Version v1

Sampling as Reasoning, Reasoning as Sampling: Structural Bridges and Their Limits Across Generative Inference

  • 1. Saluca LLC

Description

Several recent lines of work—antithetic noise in diffusion models, learned proposal distributions for sequential Monte Carlo, power-distribution sampling for reasoning elicitation, autoregressive factorization of chain-of-thought traces, and faithfulness failures in formal verification—have each been developed within their own technical communities. This paper identifies and defends a set of structural analogies connecting these lines, while carefully marking where the analogies are grounded in primary sources and where they remain conjectural. We show that antithetic noise in diffusion models is a documented instance of classical antithetic-variate variance reduction, generalized by the same authors to quasi-Monte Carlo and randomized QMC estimators. We observe that learned inference networks for SMC instantiate amortized proposal design in directed graphical models, sharing a high-level motivation with other amortized inference approaches without establishing a direct formal bridge to chain-of-thought reasoning. We note that the autoregressive factorization of chain-of-thought traces structurally resembles sequential importance-sampling decompositions, while stopping short of claiming a formal identity. Finally, we document that formal verification guarantees proof validity but not formalization faithfulness, and that specification gaming in reasoning models constitutes an empirically documented phenomenon whose decision-theoretic interpretation remains conjectural. Throughout, we distinguish grounded structural parallels from speculative analogies, providing a map of what is and is not yet established.

Authorship: Saluca Agentic AI Research Team (Saluca LLC). AI-drafted from arXiv preprint corpus on the date in the filename.

Cited arXiv preprints: 2605.30327v1

Notes

This paper was AI-drafted by an internal multi-persona research agent over a curated arXiv corpus. It is not peer-reviewed. All cited works are listed by arXiv ID; readers should follow those links to verify claims against the primary preprints.

Files

20260619_sampling-as-reasoning-reasoning-as-sampling-structural-bridg.pdf

Files (66.5 kB)