mlsanityData to 7 Oct 2026

Rewarding Efficient Reasoning Improves Abstention on Underspecified Tasks in Reasoning Models

arXiv 2609.20846 ↗First posted 5 Aug 2026cs.CL cs.LGcited by 0

On the map: LLM reasoning → Test-time compute and efficient thinking

Abstract (arXiv)

While modern large reasoning models (LRMs) excel at providing correct answers in many tasks, we provide additional evidence for the observation that they often struggle with a critical capability: knowing when to abstain from answering. We analyze this gap by comparing LRM behavior to results from a human study, revealing that human reasoning effort on unanswerable tasks is upper-bounded by answerable tasks, whereas LRMs waste computational resources by generating longer Chains of Thought (CoTs) on unanswerable than on answerable prompts. To overcome this inefficiency, we take inspiration from a resource-rational perspective on human cognition and introduce a novel GRPO reward that encourages efficient reasoning about whether the task contains all the information needed to solve it. Fine-tuning several 4B LRMs with this reward leads to human-like abstention performance gains (+12.8% on average) while retaining answering capabilities and boosting the models' efficiency (44% shorter CoTs on average).

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 32 of its 41 references that are in the map. First posted on arXiv 5 Aug 2026.

Builds on

32 of 41 references.

Closest ancestors · 9

Same direction, newest first

  1. 2026Bridging the Detection-to-Abstention Gap in Reasoning Models under Insufficient Information27 May
  2. 2026Abstain-R1: Calibrated Abstention and Post-Refusal Clarification via Verifiable RL18 Apr
  3. 2026Falcon-H1R: Pushing the Reasoning Frontiers with a Hybrid Model for Efficient Test-Time Scaling5 Jan
  4. 2025Entropy After </Think> for reasoning model early exiting30 Sep
  5. 2025Fast on the Easy, Deep on the Hard: Efficient Reasoning via Powered Length Penalty12 Jun
  6. 2025Just Enough Thinking: Efficient Reasoning with Adaptive Length Penalties Reinforcement Learning5 Jun
  7. 2025Dynamic Early Exit in Reasoning Models22 Apr
  8. 2025s1: Simple test-time scaling31 Jan
  9. 2022Language Models (Mostly) Know What They Know11 Jul

Other references · 23

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  3. 2020Measuring Massive Multitask Language Understandingcited by 7,768
  4. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,686
  5. 2023Let's Verify Step by Stepcited by 3,598
  6. 2023Sparks of Artificial General Intelligence: Early experiments with GPT-4cited by 2,579
  7. 2025Understanding R1-Zero-Like Training: A Critical Perspectivecited by 1,276
  8. 2023Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedbackcited by 788
  9. 2020ProofWriter: Generating Implications, Proofs, and Abductive Statements over Natural Languagecited by 309
  10. 2024GSM-Plus: A Comprehensive Benchmark for Evaluating the Robustness of LLMs as Mathematical Problem Solverscited by 155
  11. 2025AbstentionBench: Reasoning LLMs Fail on Unanswerable Questionscited by 129
  12. 2024Mind Your Step (by Step): Chain-of-Thought can Reduce Performance on Tasks where Thinking Makes Humans Worsecited by 107
11 more
  1. 2024STaR-GATE: Teaching Language Models to Ask Clarifying Questionscited by 99
  2. 2023The Art of Defending: A Systematic Evaluation and Analysis of LLM Defense Strategies on Safety and Over-Defensivenesscited by 63
  3. 2025QuestBench: Can LLMs ask the right question to acquire information in reasoning tasks?cited by 56
  4. 2024Benchmarking Hallucination in Large Language Models based on Unanswerable Math Word Problemcited by 42
  5. 2024Asking the Right Question at the Right Time: Human and Model Uncertainty Guidance to Ask Clarification Questionscited by 32
  6. 2024Teaching Large Language Models to Express Knowledge Boundary from Their Own Signalscited by 32
  7. 2024Don't Just Say "I don't know"! Self-aligning Large Language Models for Responding to Unknown Questions with Explanationscited by 30
  8. 2025Scaling up the think-aloud methodcited by 7
  9. 2025Beyond Passive Critical Thinking: Fostering Proactive Questioning to Enhance Human-AI Collaborationcited by 5
  10. 2026Reference Games as a Testbed for the Alignment of Model Uncertainty and Clarification Requestscited by 2
  11. 2026Act or Clarify? Modeling Sensitivity to Uncertainty and Cost in Communicationcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Answering the Unanswerable Is to Err Knowingly: Analyzing and Mitigating Abstention Failures in Large Reasoning Models
  2. 2026Thinking effort aligns between humans and reasoning models in abductive reasoning
  3. 2025When to Continue Thinking: Adaptive Thinking Mode Switching for Efficient Reasoning
  4. 2025Trade-offs in Large Reasoning Models: An Empirical Analysis of Deliberative and Adaptive Reasoning over Foundational Capabilities
  5. 2025Exploring and Exploiting the Inherent Efficiency within Large Reasoning Models for Self-Guided Efficiency Enhancement
  6. 2025Reasoning or Retrieval? A Study of Answer Attribution on Large Reasoning Models
  7. 2025The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity
  8. 2026An Enigma of Artificial Reason: Investigating the Production-Evaluation Gap in Large Reasoning Models