Whether LM question-answering consistency constitutes evidence for beliefs in a meaningful sense depends on the behaviourist operationalisation chosen.

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

• LMs can consistently exhibit more or less accurate and logically coherent beliefs on specific input distributions.claim0.829
  Conclusion from Experiment 2 on Leap-of-Thought.
• It makes little sense to speak of an LLM dialogue agent's beliefs or intentions in a literal sense, so it cannot assert a falsehood in good faith nor deliberately deceiveclaim0.800
  Philosophical claim grounding the analysis of deception in dialogue agents
• Stating and proving that answers to questions and other statements are responsive seems to require a substantially larger logical apparatus than merely proving that the answers are truthful.claim0.793
  Claim about the difficulty of responsiveness verification.
• LLMs implicitly learn a distribution of 'consistent reasoning paths', and inconsistent reasoning forms statistical outliers with low probability under this distribution.hypothesis0.789
  Theoretical hypothesis about the mechanism underlying LLM error detection and reflection.
• Some LMs can exhibit consistent intentions to be helpful and harmless, and consistently do not intend unethical instrumental goals.claim0.787
  Conclusion from Experiments 3 and 4.
• What factors determine the consistency with which LMs exhibit character traits?question0.779
  Empirical question answered through experiments on model size, fine-tuning, and prompting.
• Does the model internally maintain a form of 'consistency score' or probability mass over coherent reasoning trajectories, and how is this score modulated during reflection?question0.778
  Promising future research direction about the internal mechanism of error detection.
• We hypothesize that LLMs represent correctness of arithmetic expressions differently from factual statements.hypothesis0.777
  Core working hypothesis motivating the factual vs. arithmetic task split in the experimental design.