Me, Myself, and $π$ : Evaluating and Explaining LLM Introspection

• URL: http://arxiv.org/abs/2603.20276v1
• Date: Tue, 17 Mar 2026 17:39:25 GMT
• Title: Me, Myself, and $π$ : Evaluating and Explaining LLM Introspection
• Abstract summary: Introspection is the ability to assess and reason about one's own cognitive processes.<n>Introspect-Bench is an evaluation suite designed for rigorous capability testing.
• Score: 1.0610440409455506
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: A hallmark of human intelligence is Introspection-the ability to assess and reason about one's own cognitive processes. Introspection has emerged as a promising but contested capability in large language models (LLMs). However, current evaluations often fail to distinguish genuine meta-cognition from the mere application of general world knowledge or text-based self-simulation. In this work, we propose a principled taxonomy that formalizes introspection as the latent computation of specific operators over a model's policy and parameters. To isolate the components of generalized introspection, we present Introspect-Bench, a multifaceted evaluation suite designed for rigorous capability testing. Our results show that frontier models exhibit privileged access to their own policies, outperforming peer models in predicting their own behavior. Furthermore, we provide causal, mechanistic evidence explaining both how LLMs learn to introspect without explicit training, and how the mechanism of introspection emerges via attention diffusion.

Related papers

• CAREBench: Evaluating LLMs' Emotion Understanding by Assessing Cognitive Appraisal Reasoning [15.748287669426695]
  We introduce CAREBench, the first benchmark with complete inferential chain annotations from both first- and third-person perspectives on real-world narratives.<n>We find that stronger models match or surpass human observers on certain tasks, yet fall short on appraisal reasoning and positive emotion recognition.
  arXiv  Detail & Related papers  (2026-05-16T22:23:21Z)
• Decomposing and Steering Functional Metacognition in Large Language Models [7.62618887128252]
  We show that large language models maintain a decomposable space of functional metacognitive states.<n>We show that these states are linearly decodable from internal activations and exhibit distinct layer-wise profiles.<n>Our findings suggest that benchmark performance reflects not only task competence but also the activation of specific functional metacognitive states.
  arXiv  Detail & Related papers  (2026-05-09T13:22:47Z)
• MirrorBench: Evaluating Self-centric Intelligence in MLLMs by Introducing a Mirror [68.92922229518848]
  We introduce MirrorBench, a simulation-based benchmark inspired by the classical Mirror Self-Recognition (MSR) test in psychology.<n>MirrorBench extends this paradigm to embodied MLLMs through a tiered framework of progressively challenging tasks.<n>Our study bridges psychological paradigms and embodied intelligence, offering a principled framework for evaluating the emergence of general intelligence in large models.
  arXiv  Detail & Related papers  (2026-04-16T08:45:34Z)
• Evidence for Limited Metacognition in LLMs [2.538209532048867]
  We introduce a novel methodology for quantitatively evaluating metacognitive abilities in LLMs.<n>Taking inspiration from research on metacognition in nonhuman animals, our approach eschews model self-reports and instead tests to what degree models can strategically deploy knowledge of internal states.
  arXiv  Detail & Related papers  (2025-09-25T20:30:15Z)
• VERIFY: A Benchmark of Visual Explanation and Reasoning for Investigating Multimodal Reasoning Fidelity [34.29409506366145]
  VERIFY is a benchmark designed to isolate and rigorously evaluate the visual reasoning capabilities of state-of-the-art MLLMs.<n>Each problem is accompanied by a human-annotated reasoning path, making it the first to provide in-depth evaluation of model decision-making processes.<n>We propose novel metrics that assess visual reasoning fidelity beyond mere accuracy, highlighting critical imbalances in current model reasoning patterns.
  arXiv  Detail & Related papers  (2025-03-14T16:26:11Z)
• Evaluating Human Alignment and Model Faithfulness of LLM Rationale [66.75309523854476]
  We study how well large language models (LLMs) explain their generations through rationales. We show that prompting-based methods are less "faithful" than attribution-based explanations.
  arXiv  Detail & Related papers  (2024-06-28T20:06:30Z)
• Evaluating Large Language Models with Psychometrics [59.821829073478376]
  This paper offers a comprehensive benchmark for quantifying psychological constructs of Large Language Models (LLMs)<n>Our work identifies five key psychological constructs -- personality, values, emotional intelligence, theory of mind, and self-efficacy -- assessed through a suite of 13 datasets.<n>We uncover significant discrepancies between LLMs' self-reported traits and their response patterns in real-world scenarios, revealing complexities in their behaviors.
  arXiv  Detail & Related papers  (2024-06-25T16:09:08Z)
• MR-GSM8K: A Meta-Reasoning Benchmark for Large Language Model Evaluation [60.65820977963331]
  We introduce a novel evaluation paradigm for Large Language Models (LLMs) This paradigm shifts the emphasis from result-oriented assessments, which often neglect the reasoning process, to a more comprehensive evaluation. By applying this paradigm in the GSM8K dataset, we have developed the MR-GSM8K benchmark.
  arXiv  Detail & Related papers  (2023-12-28T15:49:43Z)
• Interpreting Pretrained Language Models via Concept Bottlenecks [55.47515772358389]
  Pretrained language models (PLMs) have made significant strides in various natural language processing tasks. The lack of interpretability due to their black-box'' nature poses challenges for responsible implementation. We propose a novel approach to interpreting PLMs by employing high-level, meaningful concepts that are easily understandable for humans.
  arXiv  Detail & Related papers  (2023-11-08T20:41:18Z)
• Learning by Self-Explaining [23.420673675343266]
  We introduce a novel workflow in the context of image classification, termed Learning by Self-Explaining (LSX) LSX utilizes aspects of self-refining AI and human-guided explanatory machine learning. Our results indicate improvements via Learning by Self-Explaining on several levels.
  arXiv  Detail & Related papers  (2023-09-15T13:41:57Z)
• Explainability for Large Language Models: A Survey [59.67574757137078]
  Large language models (LLMs) have demonstrated impressive capabilities in natural language processing. This paper introduces a taxonomy of explainability techniques and provides a structured overview of methods for explaining Transformer-based language models.
  arXiv  Detail & Related papers  (2023-09-02T22:14:26Z)
• Post Hoc Explanations of Language Models Can Improve Language Models [43.2109029463221]
  We present a novel framework, Amplifying Model Performance by Leveraging In-Context Learning with Post Hoc Explanations (AMPLIFY) We leverage post hoc explanation methods which output attribution scores (explanations) capturing the influence of each of the input features on model predictions. Our framework, AMPLIFY, leads to prediction accuracy improvements of about 10-25% over a wide range of tasks.
  arXiv  Detail & Related papers  (2023-05-19T04:46:04Z)
• Explain, Edit, and Understand: Rethinking User Study Design for Evaluating Model Explanations [97.91630330328815]
  We conduct a crowdsourcing study, where participants interact with deception detection models that have been trained to distinguish between genuine and fake hotel reviews. We observe that for a linear bag-of-words model, participants with access to the feature coefficients during training are able to cause a larger reduction in model confidence in the testing phase when compared to the no-explanation control.
  arXiv  Detail & Related papers  (2021-12-17T18:29:56Z)

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.