method
active
method:elo-preference-scoring-over-64-activitiesElo preference scoring over 64 activities
Pairwise A/B choices for 4032 pairs converted to Elo.
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.
- Scoring system used to calculate relative preference for each trait across 25,000 sampled responses and LLM-as-judge judgments
- A rating system used to compare model helpfulness and harmlessness based on crowdworker preferences.
- Pairwise comparison results converted to Elo ratings for Alexander mirror aesthetic rankings
- Best-performing run, borderline root cause.
- Primary metric for all benchmarks, measuring fraction of tasks that meet benchmark-specific pass criteria
- At n=256 best-of-n, probe scoring outperforms Gemma LLM-judge scoring by nearly 15 percentage pointsfinding0.684Test-time scaling comparison.
- More general greater-than feature.