Baseline scores blend together at least three different things: latent reflective capacity, default accessibility, and stability of access.

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

• Default behavior hides reflective capacity; models exhibit high gating between latent capacity and accessibility.finding0.809
  Grok 4: baseline 2.24, prompted 6.48; Gemini 3.1 Pro: 1.97→6.18. Reflective mode exists but is suppressed in default interaction.
• Reflective mode comprises three separable traits: latent capacity, default accessibility, and stability of access.hypothesis0.805
  Decomposition from prompt lift data: models may have capacity without accessibility (Grok 4 high-gated), and stability varies (Haiku Δ=0.02 vs GPT-5.4 Δ=1.00).
• Default presentation conflates capacity with accessibility, and most evaluation benchmarks measure only default presentation — systematically misreading models.claim0.795
  Argues current evaluation approaches are fundamentally misleading about model capabilities
• End-to-end evaluation scores conflate three sources of improvement: base capability, harness-updating quality, and harness-benefit, leaving it unclear which models produce useful updates or benefit most from themclaim0.759
  Motivating claim for the paper's controlled analysis approach
• Benchmarks of this kind test whether models can sustain strategic coherence over time, manage resource constraints, and adapt interactively — capabilities that static benchmarks do not measure.claim0.758
  Broader methodological claim about the need for multi-agent, long-horizon benchmarks.
• ReflectiveBench designed as cognitive glue (per Lyons-Levin spec) will generate higher citation velocity and lab coordination response than static-credibility benchmark.prediction0.758
• Academic consensus is shifting toward substrate neutrality of cognitive primitives.claim0.756
• We take our principal contributions in this report to be: 1. Showing that the assessment of consciousness in AI is scientifically tractable ... 2. Proposing a rubric for assessing consciousness in AI ... 3. Providing initial evidence that many of the indicator properties can be implemented in AI systems using current techniques ...quote0.754
  Summary of contributions.