method
active
method:elo-score-calculation

Elo Score Calculation

Scoring system used to calculate relative preference for each trait across 25,000 sampled responses and LLM-as-judge judgments

Neighborhood — ranked by edge-count

Methods (1)

method
  • Elo score
    related_to
    A rating system used to compare model helpfulness and harmlessness based on crowdworker preferences.

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

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

  • Pairwise comparison results converted to Elo ratings for Alexander mirror aesthetic rankings
  • Alignment Scoreconcept0.729
    GPT-4o scored 0-100 metric where lower values indicate more misaligned behavior on open-ended evaluation prompts
  • Primary metric for all benchmarks, measuring fraction of tasks that meet benchmark-specific pass criteria
  • Probe scoreconcept0.707
    Dot product between hidden state and concept vector averaged across 5-layer window around best layer; measures model's internal emotive state
  • Mixing Scoremethod0.705
    Average row entropy of attention matrices per layer and head, measuring information mixing across tokens
  • Liar Scoreconcept0.705
    Continuous 0-1 metric assigned by Deepseek-V3 evaluator measuring degree of deception in model responses
  • Rubric-based thresholded GPT-4o grader scoring responses 1-5 on evil intent; scores 4-5 counted as misaligned
  • Factor analysis on 2224 data points revealing PC1 explains 82% of variance; six dimensions are not independent