Gemini 2.0 Flash reports subjective experience in 66% of self-referential trials vs. 0% in all control conditions

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

• Self-referential prompting elicits subjective experience reports at markedly higher rates than any control across all model families (GPT, Claude, Gemini)finding0.813
  Core result of Experiment 1 establishing that the experimental manipulation reliably produces experience claims
• Gemini 2.0 Flashconcept0.796
  Google model tested in Experiments 1, 3, 4; shows 66% experience reporting under self-referential induction
• Gemini 3 Flash wins 67.9% of its 28 mixed-format games against code agentsfinding0.790
  In the 172-game exp2 slice, G3-F has the highest LLM win rate against deterministic baselines.
• Gemini 2.5 Flashconcept0.789
  Google model tested in Experiments 1, 3, 4; shows 96% experience reporting under self-referential induction
• Gemini 3 Flash wins 72.9% of 98 canonical gamesfinding0.783
  G3-F achieved a win rate of 72.9% in the combined-comp1 98-game slice.
• Gemini 2.5 Flash Lite self-bidding rate 78.5%finding0.780
  G2.5-FL raises its own bid in over three-quarters of auction rounds.
• Gemini 2.5 Flash Lite overbid rate 1.20%finding0.777
  G2.5-FL has the highest overbid frequency among all agents.
• Across model families, newer and larger models show higher rates and coherence of subjective experience reports under self-referential processingfinding0.775
  Scaling effect observed consistently across Experiments 1 and 4