mlsanityData to 14 Oct 2026

VStress: Correlation-Aware Auditing and Adaptive Budget Allocation for Repeated Verifiers

arXiv 2609.36958 ↗First posted 29 Sep 2026cs.LG cs.AI cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

Repeated verifier calls are useful only when they contribute conditional information. We introduce VStress, an auditable replay contract, and VStress-CA, a correlation-aware allocation policy that estimates the conditional marginal information of an unqueried verifier on a sealed calibration split, discounts uncertainty, normalizes by call cost, and stops or abstains when the next call is not informative. The controller freezes its decision and cost ledger before joining the clean oracle; a dependence-shift alarm disables channel preference and falls back to exact-stop. The controlled audit gives the mechanism boundary: at 35% symmetric corruption, majority-5 improves balanced accuracy from 0.6578 to 0.7739, whereas at 65% it loses 0.1226 points. In the matched fixed-budget comparison, breadth, redundancy, and adaptive allocation obtain balanced accuracies 0.6048, 0.6375, and 0.6538, with 3.4216 calls per item and an RLVR score of 0.6417 for VStress-CA. Dependence diagnostics also increase from same-model repeats to cross-family channels, with conditional marginal gains of 0.0126, 0.0462, and 0.0913. These measurements turn correlation from a post-hoc warning into an auditable allocation decision.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 40 of its 51 references that are in the map. First posted on arXiv 29 Sep 2026.

Builds on

40 of 51 references.

References · 40

  1. 2022Training language models to follow instructions with human feedbackcited by 15,971
  2. 2017Proximal Policy Optimization Algorithmscited by 15,215
  3. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  4. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  5. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
  6. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  7. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,276
  8. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,182
  9. 2023Let's Verify Step by Stepcited by 3,789
  10. 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,442
  11. 2023SWE-bench: Can Language Models Resolve Real-World GitHub Issues?cited by 3,373
  12. 2015Trust Region Policy Optimizationcited by 2,580
28 more
  1. 2023Voyager: An Open-Ended Embodied Agent with Large Language Modelscited by 2,191
  2. 2023WebArena: A Realistic Web Environment for Building Autonomous Agentscited by 2,006
  3. 2020Conservative Q-Learning for Offline Reinforcement Learningcited by 1,809
  4. 2024SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineeringcited by 1,737
  5. 2021Decision Transformer: Reinforcement Learning via Sequence Modelingcited by 1,546
  6. 2020D4RL: Datasets for Deep Data-Driven Reinforcement Learningcited by 1,427
  7. 2024τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domainscited by 1,402
  8. 2021Offline Reinforcement Learning with Implicit Q-Learningcited by 1,293
  9. 2023AgentBench: Evaluating LLMs as Agentscited by 1,267
  10. 2024OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environmentscited by 1,195
  11. 2015Prioritized Experience Replaycited by 1,099
  12. 2024Qwen2.5-Math Technical Report: Toward Mathematical Expert Model via Self-Improvementcited by 1,090
  13. 2022Planning with Diffusion for Flexible Behavior Synthesiscited by 1,061
  14. 2023FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performancecited by 853
  15. 2024RouteLLM: Learning to Route LLMs with Preference Datacited by 682
  16. 2024RewardBench: Evaluating Reward Models for Language Modelingcited by 499
  17. 2015Policy Distillationcited by 440
  18. 2019SelectiveNet: A Deep Neural Network with an Integrated Reject Optioncited by 350
  19. 2024ToolSandbox: A Stateful, Conversational, Interactive Evaluation Benchmark for LLM Tool Use Capabilitiescited by 295
  20. 2024ProcessBench: Identifying Process Errors in Mathematical Reasoningcited by 292
  21. 2017DeepStack: Expert-Level Artificial Intelligence in No-Limit Pokercited by 233
  22. 2022Defining and Characterizing Reward Hackingcited by 203
  23. 2019OpenSpiel: A Framework for Reinforcement Learning in Gamescited by 186
  24. 2016Deep Reinforcement Learning from Self-Play in Imperfect-Information Gamescited by 158
  25. 2020Combining Deep Reinforcement Learning and Search for Imperfect-Information Gamescited by 114
  26. 2018Deep Counterfactual Regret Minimizationcited by 103
  27. 2025Correlated Errors in Large Language Modelscited by 45
  28. 2026Using Common Random Numbers for Simulation-based Planning with Rolloutscited by 5

Built on by

Not cited yet.