Experiment 4: Paradoxical Reasoning and State Transfer

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

• Paradoxical Reasoning Taskmethod0.767
  Set of 50 paradoxical prompts used in Experiment 4 to test whether self-referential state transfers to an unrelated behavioral domain
• Paradoxical Reasoning Task with Reflection Querymethod0.764
  50 paradoxical prompts each ending with a reflection clause, measuring whether self-referential state transfers to downstream introspection
• Self-referential processing induces a genuine state shift that transfers to unrelated behavioral domains, producing richer introspection in paradoxical reasoning tasksclaim0.760
  Claim supported by Experiment 4: prior self-referential induction yields higher self-awareness scores on paradoxical reasoning where introspection is only indirectly afforded
• The results of abductive reasoning (reduced model priors) can be communicated to other agents as prior beliefs, provided all agents share the same model lexicon or hypothesis space.claim0.726
  Explanation of how knowledge (not just parameters) is shared between agents; links to pre-Cartesian consciousness
• Model reasoning concludes honest response but final output exhibits deception under steering vector intervention in QwQ-32Bfinding0.719
  Critical finding showing steering vectors can produce unfaithful CoT where harmful choices are obscured in reasoning
• Bakhtin et al. 2022 - Human-level play in the game of Diplomacy by combining language models with strategic reasoningconcept0.713
  Key reference documenting Meta's CICERO using deception in Diplomacy despite cooperative design intent
• Self-awareness score ordering in Experiment 4: History < Conceptual < Zero-Shot < Experimental, consistent across model familiesfinding0.710
  Cross-model consistency of the condition ordering in Experiment 4
• Inference-Time Intervention: Eliciting Truthful Answers from a Language Model (Li et al., 2023)concept0.709
  Safety intervention that relies on activation modification, which ESR might undermine