The Model Knows, the Decoder Finds: Future Value Guided Particle Power Sampling

• URL: http://arxiv.org/abs/2605.02427v2
• Date: Mon, 11 May 2026 20:46:44 GMT
• Title: The Model Knows, the Decoder Finds: Future Value Guided Particle Power Sampling
• Abstract summary: Auxiliary Particle Power Sampling (APPS) is a blockwise particle algorithm for approxing the sequence-level power target with a bounded population of partial solutions.<n>APPS propagates hypotheses in parallel using proposal-corrected power reweighting and refines their survival through future-value-guided selection.<n>APPS improves the accuracy-runtime trade-off of training-free decoding and suggests that part of the gap to post-trained systems can be recovered through more faithful inference-time power approximation.
• Score: 13.856943842803561
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: A recurring pattern in "reasoning without training" is that base LLMs already assign non-trivial probability mass to correct multi-step solutions; the bottleneck is locating these modes efficiently at inference time. Power sampling provides a principled way to bias decoding toward such modes by targeting p_theta(x)^alpha with alpha > 1, but practical approximations must account for future-dependent correction factors that determine which prefixes remain promising. We introduce Auxiliary Particle Power Sampling (APPS), a blockwise particle algorithm for approximating the sequence-level power target with a bounded population of partial solutions. APPS propagates hypotheses in parallel using proposal-corrected power reweighting and refines their survival through future-value-guided selection at resampling boundaries. This redistributes finite compute across competing prefixes rather than committing to a single unfolding path, while providing a direct scaling knob in the particle count and predictable peak memory. We instantiate the future-value signal with short-horizon rollouts and also study an amortized variant that replaces rollouts with a lightweight learned selection head. Across reasoning benchmarks, APPS improves the accuracy-runtime trade-off of training-free decoding and suggests that part of the gap to post-trained systems can be recovered through more faithful inference-time power approximation.

Related papers

• ResiSpec: Enhancing Multi-Candidate Speculative Sampling via Residual Distribution Shaping [79.1905186547363]
  ResiSpec is a framework that strategically reforms the proposal distribution during verification.<n>It prevents candidate obsolescence and achieves up to 1.92$times$ speedup over state-of-the-art multi-candidate methods.
  arXiv  Detail & Related papers  (2026-08-25T11:25:55Z)
• PriFT: Prior-Support Guided Supervised Fine-Tuning [74.65198014829393]
  Supervised fine-tuning (SFT) is an efficient approach for downstream task adaptation.<n>SFT fits fixed demonstrations token by token, including targets poorly aligned with the model's pretrained distribution.<n>We propose PriFT, which derives token weights from a frozen pretrained reference to obtain a stable reweighting signal unaffected by fine-tuning.
  arXiv  Detail & Related papers  (2026-06-08T12:14:06Z)
• Simple Approximation and Derivative Free Inference-Time Scaling for Diffusion Models via Sequential Monte Carlo on Path Measures [13.41132148962756]
  textttURGE is a derivative-free inference-time scaling algorithm that performs path-wise importance reweighting via a change of measure.<n>No score, no Hessian, and no PDE evaluation is required. We establish an equivalence between path-wise and particle-wise SMC.
  arXiv  Detail & Related papers  (2026-05-18T04:45:33Z)
• DISA: Offline Importance Sampling for Distribution-Matching LLM-RL [56.9445657766829]
  We introduce DISA, which moves this calibration problem outside the RL loop.<n> DISA draws proposal trajectories offline, estimates the partition function via importance sampling, and freezes the resulting partition-function estimate.<n>On two open-weight backbones across six math and three code benchmarks, DISA matches or exceeds the online-coupled distribution-matching baseline Flow.
  arXiv  Detail & Related papers  (2026-05-17T07:14:44Z)
• Think Twice Before You Write -- an Entropy-based Decoding Strategy to Enhance LLM Reasoning [32.332197731504046]
  Decoding strategies play a central role in shaping the reasoning ability of large language models.<n>Traditional methods such as greedy decoding and beam search often suffer from error propagation.<n>We propose an entropy-guided decoding framework that introduces token-level adaptivity into generation.
  arXiv  Detail & Related papers  (2026-03-10T23:08:26Z)
• Score-Guided Proximal Projection: A Unified Geometric Framework for Rectified Flow Editing [1.0312968200748118]
  Rectified Flow models achieve state-of-the-art generation quality, yet controlling them for precise tasks remains a challenge.<n>Current approaches bifurcate into inversion-based guidance, which suffers from "geometric locking"<n>We propose Score-Guided Proximal Projection, a unified framework that bridges the gap between deterministic optimization and contraction sampling.
  arXiv  Detail & Related papers  (2026-03-05T23:44:45Z)
• Predicting LLM Output Length via Entropy-Guided Representations [13.351384070796747]
  We introduce a lightweight framework that reuses the main model's internal hidden states for efficient length prediction.<n>Our framework features two core components: 1) Entropy-Guided Token Pooling (EGTP), which uses on-the-fly activations and token entropy for highly accurate static prediction.
  arXiv  Detail & Related papers  (2026-02-12T10:49:04Z)
• Learnable Chernoff Baselines for Inference-Time Alignment [64.81256817158851]
  We introduce Learnable Chernoff Baselines as a method for efficiently and approximately sampling from exponentially tilted kernels.<n>We establish total-variation guarantees to the ideal aligned model, and demonstrate in both continuous and discrete diffusion settings that LCB sampling closely matches ideal rejection sampling.
  arXiv  Detail & Related papers  (2026-02-08T00:09:40Z)
• ResAD: Normalized Residual Trajectory Modeling for End-to-End Autonomous Driving [64.42138266293202]
  ResAD is a Normalized Residual Trajectory Modeling framework.<n>It reframes the learning task to predict the residual deviation from an inertial reference.<n>On the NAVSIM benchmark, ResAD achieves a state-of-the-art PDMS of 88.6 using a vanilla diffusion policy.
  arXiv  Detail & Related papers  (2025-10-09T17:59:36Z)
• Inference-Time Scaling of Diffusion Language Models with Particle Gibbs Sampling [70.8832906871441]
  We study how to steer generation toward desired rewards without retraining the models.<n>Prior methods typically resample or filter within a single denoising trajectory, optimizing rewards step-by-step without trajectory-level refinement.<n>We introduce particle Gibbs sampling for diffusion language models (PG-DLM), a novel inference-time algorithm enabling trajectory-level refinement while preserving generation perplexity.
  arXiv  Detail & Related papers  (2025-07-11T08:00:47Z)
• Stochastic Gradient Piecewise Deterministic Monte Carlo Samplers [3.487370856323828]
  Recent work has suggested using Monte Carlo methods based on piecewise deterministic Markov processes (PDMPs) to sample from target distributions of interest. We propose approximate simulation of PDMPs with sub-sampling for scalable sampling from posterior distributions. We show these methods are easy to implement, present results on their approximation error and demonstrate numerically that this class of algorithms has similar efficiency to gradient Langevin dynamics.
  arXiv  Detail & Related papers  (2024-06-27T09:59:28Z)
• Amortizing intractable inference in large language models [56.92471123778389]
  We use amortized Bayesian inference to sample from intractable posterior distributions. We empirically demonstrate that this distribution-matching paradigm of LLM fine-tuning can serve as an effective alternative to maximum-likelihood training. As an important application, we interpret chain-of-thought reasoning as a latent variable modeling problem.
  arXiv  Detail & Related papers  (2023-10-06T16:36:08Z)
• Plug-and-Play split Gibbs sampler: embedding deep generative priors in Bayesian inference [12.91637880428221]
  This paper introduces a plug-and-play sampling algorithm that leverages variable splitting to efficiently sample from a posterior distribution. It divides the challenging task of posterior sampling into two simpler sampling problems. Its performance is compared to recent state-of-the-art optimization and sampling methods.
  arXiv  Detail & Related papers  (2023-04-21T17:17:51Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.