mlsanityData to 14 Oct 2026

RCCM: Diagnosing Gains and Harms in Reasoning-Prompt Transfer

arXiv 2610.19872 ↗First posted 14 Oct 2026cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

A reasoning prompt can improve one deployment and degrade another. RCCM (Reasoning-Control Capability Mismatch) examines these reversals by pairing answers to the same questions with records of how generation ends. Across five open-weight deployments and nine strategies, aggregate accuracy hides substantial question-level gains and harms; a common-budget MATH500 evaluation examines transfer on a separate sample. On 60 additional questions, concise reasoning improves Phi's automatic answer matching at a 2,048-token cap across three collections, but sharply reduces Qwen's. Author-reviewed final-answer commitments give a smaller, statistically unresolved Phi gain. Reserving an answer stage also improves matching, with additional input and call costs. These results show why a prompt benefit must be interpreted alongside its execution and evaluation: matching a reference answer need not mean delivering it, and an apparent compression benefit may admit another explanation. RCCM records this evidence for reproducible transfer comparisons.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 25 of its 39 references that are in the map. First posted on arXiv 14 Oct 2026.

Builds on

25 of 39 references.

References · 25

  1. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  2. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,371
  3. 2023Let's Verify Step by Stepcited by 3,789
  4. 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,442
  5. 2023Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Promptingcited by 1,122
  6. 2023Large Language Models Cannot Self-Correct Reasoning Yetcited by 1,017
  7. 2023Quantifying Language Models' Sensitivity to Spurious Features in Prompt Design or: How I learned to start worrying about prompt formattingcited by 817
  8. 2023State of What Art? A Call for Multi-Prompt LLM Evaluationcited by 310
  9. 2025Chain of Draft: Thinking Faster by Writing Lesscited by 297
  10. 2024Chain-of-Thought Reasoning Without Promptingcited by 210
  11. 2024The Impact of Reasoning Step Length on Large Language Modelscited by 153
  12. 2024ProSA: Assessing and Understanding the Prompt Sensitivity of LLMscited by 149
13 more
  1. 2025Sketch-of-Thought: Efficient LLM Reasoning with Adaptive Cognitive-Inspired Sketchingcited by 117
  2. 2025How Well do LLMs Compress Their Own Chain-of-Thought? A Token Complexity Approachcited by 97
  3. 2024Reasoning in Token Economies: Budget-Aware Evaluation of LLM Reasoning Strategiescited by 45
  4. 2023How are Prompts Different in Terms of Sensitivity?cited by 35
  5. 2025Steering LLM Thinking with Budget Guidancecited by 33
  6. 2025Flaw or Artifact? Rethinking Prompt Sensitivity in Evaluating LLMscited by 30
  7. 2025Right Answer, Wrong Score: Uncovering the Inconsistencies of LLM Evaluation in Multiple-Choice Question Answeringcited by 28
  8. 2024Instance-adaptive Zero-shot Chain-of-Thought Promptingcited by 26
  9. 2025An Empirical Study of LLM Reasoning Ability Under Strict Output Length Constraintcited by 17
  10. 2026When More Thinking Hurts: Overthinking in LLM Test-Time Compute Scalingcited by 16
  11. 2026Mid-Think: Training-Free Intermediate-Budget Reasoning via Token-Level Triggerscited by 6
  12. 2026Beyond the Answer: Decoding the Behavior of LLMs as Scientific Reasonerscited by 2
  13. 2025Format-Adapter: Improving Reasoning Capability of LLMs by Adapting Suitable Formatcited by 1

Built on by

Not cited yet.