Verbose reasoning chains are not required for strong play.

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

• Performative chain-of-thought is real; verbalized output does not equal internal state.claim0.763
• does chain-of-thought text faithfully reveal a model's internal reasoning process, or does it constitute performative theater?question0.746
  Central research question motivating the paper
• Chain-of-thought reasoning improves the transparency and performance of AI decision making in harmlessness evaluation.claim0.730
  CoT improves accuracy on HHH evals and makes the decision process legible.
• 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.claim0.728
  Claim about the difficulty of responsiveness verification.
• A small number of high-quality human demonstrations of chain-of-thought reasoning could be used to improve and focus performance.hypothesis0.726
  Section 6 mentions high-quality human demos could improve natural language feedback.
• Commonsense reasoning shows uniform but weaker anchoring (S ≈ −2.15)finding0.726
  Task-specific comparison.
• Chain-of-thought prompting elicits reasoning in large language models (Wei et al., 2022)concept0.724
  Foundational paper on CoT prompting cited as basis for reasoning LLM training
• Complement syntax and mental state verb comprehension abilities crucial for human ToM development are not significantly represented in LLMs, revealing fundamental discrepancies between natural and artificial intelligence regarding mind development.claim0.721
  Derived from the finding that linguistic span focusing on complements/MSV yields no significant IIT estimate changes.