Verbalized probabilities for programming languages show only Pearson r=0.182 (GPT-4.1) correlation with corpus frequencies, indicating weaker calibration

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

• GPT-4.1 verbalized probabilities for US states show Pearson r=0.741, Spearman ρ=0.896 correlation with corpus-based reference frequenciesfinding0.874
  Indicates verbalized probabilities contain meaningful distributional information for constrained answer spaces
• Probes trained on the likely dataset perform worse than chance on datasets with anti-correlations between text probability and truthfinding0.771
  Shows that truth representations are not reducible to text probability representations
• Sycophancy persona vector decomposes into stylistic features: affirmative phrases (F121978, TES=98.6), advertising language (F86620, TES=88.0), and encouraging language (F39874, TES=87.2)finding0.760
  SAE analysis shows sycophancy is primarily stylistic rather than content-based
• Confidence NLI Diversity achieves ρ=0.64 correlation with human diversity judgments on conTestfinding0.759
  Highest human correlation for semantic diversity metric
• Under reward shaping (G=100, H=-100, F=0), Active Inference scored 99.52, Bayesian RL 99.77, Q-learning 95.56, with nearly identical behavior between belief-based agents.finding0.755
  Table 2, row 3, showing equivalence when prior preferences match rewards.
• We hypothesize that native self-report, fine-tuned introspection models, and trained activation-to-language systems will show different performance on bias-resistant localization and strength benchmarkshypothesis0.753
  Comparative prediction motivating future work contrasting different approaches to LLM self-knowledge
• Verbal reports (the Turing Test) and homology to human brains are utterly inadequate criteria for assessing the status of novel, unconventional agents that offer no familiar touchstone of phylogeny or anatomy.claim0.752
  Core claim that standard criteria fail for novel agents.
• Binary detection accuracy (up to 97.3% at L0 α=5) is entirely explained by global logit shifts (r=0.999 correlation with control)finding0.751
  Core negative result: the binary detection paradigm cannot distinguish genuine introspection from uniform output bias