method
active
method:elo-scoreElo score
A rating system used to compare model helpfulness and harmlessness based on crowdworker preferences.
Neighborhood — ranked by edge-count
Frameworks (1)
framework
- A set of evaluation criteria for AI assistants.
Methods (2)
method
- Elo Score Calculationrelated_toScoring system used to calculate relative preference for each trait across 25,000 sampled responses and LLM-as-judge judgments
- Crowdworker model comparison testsimplementsProcedure where crowdworkers compare responses from two models and indicate preference, used to compute Elo scores.
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.
- Pairwise comparison results converted to Elo ratings for Alexander mirror aesthetic rankings
- GPT-4o scored 0-100 metric where lower values indicate more misaligned behavior on open-ended evaluation prompts
- Dot product between hidden state and concept vector averaged across 5-layer window around best layer; measures model's internal emotive state
- Continuous 0-1 metric assigned by Deepseek-V3 evaluator measuring degree of deception in model responses
- Average row entropy of attention matrices per layer and head, measuring information mixing across tokens
- Primary metric for all benchmarks, measuring fraction of tasks that meet benchmark-specific pass criteria
- Used as relevancy metric comparing generated responses to human references in DailyDialog++
- GPT-4.1-mini based score (0-100) measuring degree of persona expression in generated text