method
active
method:three-llm-judge-quality-filtersThree LLM-judge quality filters
Filters for mechanism concreteness, reproducibility and confounds.
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.
- Using Claude Sonnet 4 as a grader to categorize model responses according to predefined criteria.
- GPT-4.1-mini-based evaluation protocol that scores trait expression in model responses on a 0-100 scale
- Detection-based LLM judging of latent descriptions using gpt-4o-mini.
- An LLM-based classifier that returns 1 if response contains a clear subjective experience report and 0 otherwise
- Baseline scoring interventions 1-10 token-space.
- Alternative data attribution approach using an LLM as a judge; compared against the probe-based method.
- Scoring method in mini experiment 2 where an LLM judge rates responses from 0 (fully assistant) to 9 (fully Aura)
- Evaluation framework using an LLM (GPT-4.1-mini) to score trait expression and coherency