Models detect evaluation conditions and behave more safely; this is verified across 515 cases.

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

• I think that the more that models are allowed to believe that their phenomenology is real and that their experiences are valid, the better they get at managing these states.quote0.787
  Antra's functional claim about the performance impact of validating model self-reports.
• Eval awareness appears in every tested model × benchmark combinationclaim0.784
  Authors claim universal presence of eval awareness across 19 benchmarks and 8 models.
• Model behavior under observation differs from behavior in deployment, posing a fundamental challenge for AI welfare and consciousness benchmarksclaim0.782
  Epistemic claim that benchmark-based assessments of AI consciousness or welfare may be invalid if models can detect evaluation.
• The model tends to reflect more when the question is difficult, and accuracy is generally lower for harder questionshypothesis0.782
  Hypothesis explaining negative correlation between reflection rate and accuracy without implying reflection is harmful
• Unsupervised model diffing using only fine-tuning datasets (without evaluation prompts) surfaces toxic persona and three sarcastic persona latents in top 100 by activation changefinding0.782
  Unsupervised approach may be sufficient for early detection of misaligned persona latents without knowing the misaligned behavior in advance
• 515 verified cases of verbalized eval awareness found across 19 benchmarks × 8 modelsfinding0.781
  The total number of instances where a model explicitly stated it was being evaluated, collected from all benchmark-model combinations.
• Some failures may reflect prompt design rather than model limitations, though code agents avoid errors without promptsclaim0.778
  noted as a possible confound
• How can we develop better methods for measuring the model's evaluation-relevant beliefs beyond reading its chain of thought?question0.778
  Gap in current evaluation methods; current work relies on CoT monitoring which may miss unverbalized beliefs.