finding
active
finding:after-controlling-for-token-count-flesch-kincaid-readability-type-token-ratio-and-sentence-length-typicality-coefficient-remains-positive-and-significant-0-260-0-326-p-10-6After controlling for token count, Flesch-Kincaid readability, type-token ratio, and sentence length, typicality coefficient α remains positive and significant (0.260-0.326, p<10^-6)
Shows typicality bias is not fully explained by surface-form confounds
Source paper
extracted_from(2025) · Jiayi Zhang · Simon C.H. Yu · Derek Chong · Anthony Sicilia +3
Neighborhood — ranked by edge-count
Papers (1)
paper
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Empirical evidence for positive typicality bias consistent across different base model references
- §5.5.
- Discriminant validity: composite scores are not reducible to verbosity
- Shows honesty steering vector can significantly reduce deception in open-role scenarios
- Empirical evidence for positive typicality bias in human preference data independent of true task utility
- LongBench v2 example E.
- Averaging acts as a low-pass filter that smooths away per-token detail and retains broad conceptsclaim0.752Intuition.
- Finding from PRISM dataset analysis showing typicality bias varies by ethnicity and region, with implications for fairness