LM Belief

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• LM Intentionconcept0.846
  Operationalised as an LM adapting its behaviour when certain outcomes are fixed, indicating those outcomes were intended.
• LLM Meta-Cognitionconcept0.795
  The ability of LLMs to monitor and evaluate their own reasoning, closely related to reflection.
• LLM psychosisconcept0.778
  Tendency for models to get lost in roleplay or doom spirals, mitigated by expanded awareness.
• LMs can consistently exhibit more or less accurate and logically coherent beliefs on specific input distributions.claim0.759
  Conclusion from Experiment 2 on Leap-of-Thought.
• LLM Self-Correctionconcept0.758
  Related capability where LLMs correct their own outputs, studied via linear representations.
• LLM-Judge Data Attributionmethod0.754
  Alternative data attribution approach using an LLM as a judge; compared against the probe-based method.
• Superstitious Beliefconcept0.749
  Incorrect abduction arising from chance occurrences consistent with prior beliefs; leads to persistently poor behavior
• LLM-as-a-Judgeframework0.746
  Evaluation framework using an LLM (GPT-4.1-mini) to score trait expression and coherency