Capability Self-Assessment: Teaching LLMs to Know Their Limits

• URL: http://arxiv.org/abs/2606.00251v1
• Date: Fri, 29 May 2026 18:32:14 GMT
• Title: Capability Self-Assessment: Teaching LLMs to Know Their Limits
• Abstract summary: We show that modern large language models systematically lack the ability to self-assessment.<n>We formulate it as a policy-learning problem, aiming to improve self-assessment while preserving the model's original capabilities.<n>Our results show that reinforcement learning teaches CSA effectively, significantly outperforming supervised fine-tuning.
• Score: 54.08927435222828
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The ability to recognize one's own limitations and decide whether to solve a problem or delegate is fundamental for reliable intelligent systems. Yet we show that modern large language models systematically lack this ability: across diverse model families and scales, they overestimate their competence and attempt queries they cannot solve. We refer to this ability as Capability Self-Assessment (CSA) and formulate it as a policy-learning problem, aiming to improve self-assessment while preserving the model's original capabilities. Our results show that reinforcement learning teaches CSA effectively, significantly outperforming supervised fine-tuning while preserving original capabilities. In contrast, supervised fine-tuning severely degrades the capabilities the model is meant to assess. Moreover, learned self-assessment behavior generalizes well out of distribution, suggesting that CSA is a transferable model trait. Finally, CSA is practically useful: it improves local-cloud decision making at inference time and provides a signal for targeted data selection during training.

Related papers

• Evaluating and Improving LLM Self-Modeling [1.844645605886669]
  We focus on verifiable behavioral questions, such as whether a prompt edit would change the model's final answer.<n>Current models show non-trivial but limited self-modeling skill.<n>We develop a scalable synthetic-data pipeline that produces self-modeling training data.
  arXiv  Detail & Related papers  (2026-08-31T15:37:51Z)
• Reinforcement Learning with Metacognitive Feedback Elicits Faithful Uncertainty Expression in LLMs [60.80435138270317]
  We apply reinforcement learning with metacognitive feedback to the problem of faithful calibration.<n>Experiments show RLMF achieves generalizable, state-of-the-art FC on diverse tasks.<n> RLMF surpasses standard RL by up to 63% while enhancing models' ability to assess and express their own capability limits.
  arXiv  Detail & Related papers  (2026-06-30T17:56:01Z)
• The Path of Self-Evolving Large Language Models: Achieving Data-Efficient Learning via Intrinsic Feedback [51.144727949988436]
  Reinforcement learning (RL) has demonstrated potential to enhance the reasoning capabilities of large language models (LLMs)<n>In this work, we explore improving LLMs through RL with minimal data.<n>To minimize data dependency, we introduce two novel mechanisms grounded in self-awareness.
  arXiv  Detail & Related papers  (2025-10-03T06:32:10Z)
• Uncovering Intrinsic Capabilities: A Paradigm for Data Curation in Vision-Language Models [26.897329298659717]
  We introduce Capability-Attributed Data Curation (CADC), a framework that shifts from task-specific curations to intrinsic capability analysis.<n>CADC discovers intrinsic capabilities in an unsupervised manner from gradient-based learning trajectories.<n>With as little as 5% of the original data, CADC surpasses full-data training on multimodal benchmarks.
  arXiv  Detail & Related papers  (2025-09-27T02:57:37Z)
• Autonomous Learning From Success and Failure: Goal-Conditioned Supervised Learning with Negative Feedback [2.36462256498849]
  Goal-Conditioned Supervised Learning has emerged as a potential solution by enabling self-imitation learning for autonomous systems.<n>We propose a novel model that integrates contrastive learning principles into the GCSL framework to learn from both success and failure.
  arXiv  Detail & Related papers  (2025-09-03T10:50:48Z)
• TopK Language Models [23.574227495324568]
  TopK LMs offer a favorable trade-off between model size, computational efficiency, and interpretability.<n>These features make TopK LMs stable and reliable tools for understanding how language models learn and represent concepts.
  arXiv  Detail & Related papers  (2025-06-26T16:56:43Z)
• Can Large Reasoning Models Self-Train? [51.0277533541394]
  We use majority voting as a simple self-feedback mechanism to study whether self-training can be sustained within reinforcement learning.<n>We find that this basic approach improves not only the model's reasoning performance, but also its capability of generating better quality feedback for the next RL iteration.<n>Yet our analysis also reveals a critical limitation of such a self-training paradigm - prolonged RL with self-reward leads to reward hacking, resulting in sudden and complete performance collapse.
  arXiv  Detail & Related papers  (2025-05-27T17:16:00Z)
• ReVISE: Learning to Refine at Test-Time via Intrinsic Self-Verification [53.80183105328448]
  Refine via Intrinsic Self-Verification (ReVISE) is an efficient framework that enables LLMs to self-correct their outputs through self-verification.<n>Our experiments on various reasoning tasks demonstrate that ReVISE achieves efficient self-correction and significantly improves reasoning performance.
  arXiv  Detail & Related papers  (2025-02-20T13:50:02Z)
• Self-Improvement in Language Models: The Sharpening Mechanism [70.9248553790022]
  We offer a new perspective on the capabilities of self-improvement through a lens we refer to as sharpening.<n>Motivated by the observation that language models are often better at verifying response quality than they are at generating correct responses, we formalize self-improvement as using the model itself as a verifier during post-training.<n>We analyze two natural families of self-improvement algorithms based on SFT and RLHF.
  arXiv  Detail & Related papers  (2024-12-02T20:24:17Z)
• On the Intrinsic Self-Correction Capability of LLMs: Uncertainty and Latent Concept [36.27550578296276]
  Large Language Models (LLMs) are able to improve their responses when instructed to do so, a capability known as self-correction. In intrinsic self-correction is evident in various applications, but how and why it is effective remains unknown. We show that intrinsic self-correction can be progressively improved, allowing it to approach a converged state.
  arXiv  Detail & Related papers  (2024-06-04T14:55:43Z)
• Maximum Likelihood Distillation for Robust Modulation Classification [50.51144496609274]
  We build on knowledge distillation ideas and adversarial training to build more robust AMC systems. We propose to use the Maximum Likelihood function, which could solve the AMC problem in offline settings, to generate better training labels.
  arXiv  Detail & Related papers  (2022-11-01T21:06:11Z)

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.