mlsanityData to 7 Oct 2026

Emergent Introspection in AI is Content-Agnostic

arXiv 2603.05414 ↗First posted 5 Mar 2026cs.AI cs.CLcited by 5

Not on the map. Why?

Abstract (arXiv)

Introspection is a foundational cognitive ability, but its mechanism is not well understood. Recent work has shown that AI models can introspect. We study the mechanism of this introspection. We first extensively replicate Lindsey (2025)'s thought injection detection paradigm in large open-source models. We show that introspection in these models is content-agnostic: models can detect that an anomaly occurred even when they cannot reliably identify its content. The models confabulate injected concepts that are high-frequency and concrete (e.g., "apple"). They also require fewer tokens to detect an injection than to guess the correct concept (with wrong guesses coming earlier). We argue that a content-agnostic introspective mechanism is consistent with leading theories in philosophy and psychology.

Key facts. Cited by 5 papers in the map (data through 7 Oct 2026). Builds on 12 of its 25 references that are in the map. First posted on arXiv 5 Mar 2026.

Builds on

12 of 25 references.

References · 12

  1. 2025Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMscited by 158
  2. 2024Looking Inward: Language Models Can Learn About Themselves by Introspectioncited by 105
  3. 2025Tell me about yourself: LLMs are aware of their learned behaviorscited by 86
  4. 2026Emergent Introspective Awareness in Large Language Modelscited by 73
  5. 2023Could a Large Language Model be Conscious?cited by 70
  6. 2025Language Models Fail to Introspect About Their Knowledge of Languagecited by 28
  7. 2025Training Language Models to Explain Their Own Computationscited by 27
  8. 2025Self-Interpretability: LLMs Can Describe Complex Internal Processes that Drive Their Decisionscited by 23
  9. 2025Does It Make Sense to Speak of Introspection in Large Language Models?cited by 22
  10. 2025Privileged Self-Access Matters for Introspection in AIcited by 20
  11. 2026Latent Introspection: Models Can Detect Prior Concept Injectionscited by 12
  12. 2025Steering Awareness: Detecting Activation Steering from Withincited by 7

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 5.

5 papers.

  1. 2026Introspective Coupling: Self-Explanation Training Tracks Behavioral Change Despite Fixed Supervisioncited by 3
  2. 2026Phase Transitions in Driven Informational Systems: A Two-Field Perspective on Learning Theory and Non-Equilibrium Chemistrycited by 0
  3. 2026Instrument Effects in Language-Model Honesty Evaluation: An Auditable Single-System Demonstrationcited by 0
  4. 2026Open-Weight Masked Introspection: Measuring What Language Models Can Report About Their Own Computationcited by 0
  5. 2026Measuring Activation Control in Large Language Modelscited by 0