AutoCRAT: Within-trajectory Joint Control of Stochasticity and Compute for LLM Reasoning

• URL: http://arxiv.org/abs/2608.29988v1
• Date: Sun, 30 Aug 2026 19:24:58 GMT
• Title: AutoCRAT: Within-trajectory Joint Control of Stochasticity and Compute for LLM Reasoning
• Abstract summary: Large language models (LLMs) achieve strong reasoning performance, which depends critically on time decisions.<n>Recent adaptive methods partially address this limitation, but they primarily adapt either recommended decodingity (how the model explores) or reasoning compute (how long the model reasons) in isolation.<n>We introduce a within-trajectory joint control view, and instantiate it in AutoCRAT.
• Score: 10.408251801653723
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language models (LLMs) achieve strong reasoning performance, which depends critically on inference-time decisions. Yet these decisions are commonly handled by static, one-size-fits-all policies, limiting adaptation to diverse tasks and reasoning stages. Recent adaptive methods partially address this limitation, but they primarily adapt either decoding stochasticity (how the model explores) or reasoning compute (how long the model reasons) in isolation, leaving their interaction within a single reasoning trajectory unmodeled. To address this challenge, we shift toward a within-trajectory joint control view, and instantiate it in AutoCRAT, a decoder-side controller for frozen backbones. Using only signals available during decoding, AutoCRAT jointly adjusts sampling stochasticity and reasoning budget during generation. AutoCRAT operates over a discrete action space and updates control decisions only at semantic boundaries, improving stability while remaining responsive to the evolving reasoning process. Comprehensive evaluation across 6 benchmarks demonstrates that AutoCRAT (I) uses 13.8-52.7% fewer inference tokens on average than recommended static configurations, (II) surpasses recommended static and adaptive baselines by 1.5-4.5% in relative accuracy, and (III) enjoys strong cross-backbone transferability.

Related papers

• SVR: Self-Verifying Refinement via Joint Verdict-Confidence Reinforcement Learning for Adaptive Test-Time Compute [62.3458279176813]
  Self-Verifying Refinement is an oracle-free multi-turn reinforcement learning framework.<n>It learns to use self-verification as a compute-control policy.<n>It achieves a macro-average accuracy of 0.563 with only 2.99 inference turns on average.
  arXiv  Detail & Related papers  (2026-07-30T16:20:58Z)
• CLEAR: Cognition and Latent Evaluation for Adaptive Routing in End-to-End Autonomous Driving [19.474428775260034]
  We propose CLEAR, a framework that combines ultra-fast generative planning with deep semantic reasoning.<n> CLEAR employs Drive-JEPA as the visual encoder and replaces the multi-step denoising chain with a single-step conditional drift in a VAE latent space.<n>On the NAVSIM v1 benchmark, CLEAR achieves a state-of-the-art PDMS of 93.7.
  arXiv  Detail & Related papers  (2026-06-04T14:32:10Z)
• CANMOT: Class-Aware Noise Modeling for Multi-Object Tracking in Autonomous Driving [1.2744523252873352]
  CANMOT is a class-aware and object-aligned noise modeling framework for KF-based 3D MOT.<n>Experiments show that class-aware and object-aligned noise modeling improves tracking performance and substantially reduces identity.<n>Results reveal severe overconfidence in standard KF-based MOT baselines.
  arXiv  Detail & Related papers  (2026-06-02T12:56:31Z)
• VLA-ATTC: Adaptive Test-Time Compute for VLA Models with Relative Action Critic Model [54.35791816657227]
  We introduce textbfVLA-ATTC', a framework that endows VLA models with adaptive test-time compute.<n>VLA-ATTC employs an uncertainty-based cognitive clutch'' to dynamically transition from reflexive execution to a TTC deliberation phase.<n>On the LIBERO-LONG benchmark, VLA-ATTC reduces the failure rate of the SOTA model PI0.5 by over 50%.
  arXiv  Detail & Related papers  (2026-05-02T02:13:11Z)
• AdaptEvolve: Improving Efficiency of Evolutionary AI Agents through Adaptive Model Selection [14.17960333915609]
  Evolutionary agentic systems intensify the trade-off between computational efficiency and reasoning capability.<n>We introduce AdaptEvolve: Adaptive Selection for Multi-LLM Evolutionary Refinement.
  arXiv  Detail & Related papers  (2026-02-12T13:26:56Z)
• Internalizing LLM Reasoning via Discovery and Replay of Latent Actions [4.830503861275364]
  Internalization of chain-of-thought processes into hidden states has emerged as a highly efficient paradigm for scaling test-time compute.<n>We propose STIR (Self-Distilled Tools for Internal Reasoning), a framework that reformulates reasoning enhancement as a dynamic latent trajectory control problem.
  arXiv  Detail & Related papers  (2026-02-04T08:44:57Z)
• Not All Preferences Are Created Equal: Stability-Aware and Gradient-Efficient Alignment for Reasoning Models [52.48582333951919]
  We propose a dynamic framework designed to enhance alignment reliability by maximizing the Signal-to-Noise Ratio of policy updates.<n>SAGE (Stability-Aware Gradient Efficiency) integrates a coarse-grained curriculum mechanism that refreshes candidate pools based on model competence.<n> Experiments on multiple mathematical reasoning benchmarks demonstrate that SAGE significantly accelerates convergence and outperforms static baselines.
  arXiv  Detail & Related papers  (2026-02-01T12:56:10Z)
• RISER: Orchestrating Latent Reasoning Skills for Adaptive Activation Steering [62.63376387138257]
  We propose a plug-and-play intervention framework that adaptively steers large language models (LLMs) reasoning in activation space.<n>RISER constructs a library of reusable reasoning vectors and employs a lightweight Router to dynamically compose them for each input.<n>The Router is optimized via reinforcement learning under task-level rewards, activating latent cognitive primitives in an emergent and compositional manner.
  arXiv  Detail & Related papers  (2026-01-14T08:04:33Z)
• e1: Learning Adaptive Control of Reasoning Effort [88.51897900019485]
  Increasing the thinking budget of AI models can significantly improve accuracy, but not all questions warrant the same amount of reasoning.<n>Users may prefer to allocate different amounts of reasoning effort depending on how they value output quality versus latency and cost.<n>We propose Adaptive Effort Control, a self-adaptive reinforcement learning method that trains models to use a user-specified fraction of tokens.
  arXiv  Detail & Related papers  (2025-10-30T23:12:21Z)
• Alpamayo-R1: Bridging Reasoning and Action Prediction for Generalizable Autonomous Driving in the Long Tail [85.47497935739936]
  Alpamayo-R1 (AR1) is a vision-language-action model that integrates Chain of Causation reasoning with trajectory planning.<n>We show AR1 achieves 12% improvement in planning accuracy on challenging cases compared to a trajectory-only baseline.<n>We plan to release AR1 models and a subset of the CoC in a future update.
  arXiv  Detail & Related papers  (2025-10-30T01:25:34Z)
• Incorporating Recurrent Reinforcement Learning into Model Predictive Control for Adaptive Control in Autonomous Driving [11.67417895998434]
  Model Predictive Control (MPC) is attracting tremendous attention in the autonomous driving task as a powerful control technique. In this paper, we reformulate the problem as a Partially Observed Markov Decision Process (POMDP) We then learn a recurrent policy continually adapting the parameters of the dynamics model via Recurrent Reinforcement Learning (RRL) for optimal and adaptive control.
  arXiv  Detail & Related papers  (2023-01-30T22:11:07Z)

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.