Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Efficient Bayesian Inference from Noisy Pairwise Comparisons
Till Aczel, Lucas Theis, Roger Wattenhofer
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add rater-quality modeling via EM (BBQ vs Bayes-BT) | BT aggregation · IHQ-unscreened | Top-1 accuracy [%]: 23.59 -> 61.42 | Improved |
| replace gradient descent with EM (BBQ vs Crowd-BT) | rater-quality BT · IHQ-all | Top-1 accuracy [%]: 85.46 -> 99.34 | Improved |
| add rater-quality modeling on homogeneous high-quality raters | BBQ · IHQ-screened, WD | Kendall's tau: 0.9204 vs 0.9238 (IHQ-screened, Crowd-BT) | No clear effect |
| screen raters for visual checks | all BT variants · IHQ | Top-1 accuracy [%]: 61.42 -> 99.86 (BBQ, unscreened -> screened) | Improved |
| use closed-form EM instead of Crowd-BT gradient descent | BBQ · HUMAINE | runtime: ~15 minutes (Crowd-BT) vs a few seconds | Improved |
| increase number of raters or comparisons per rater | BBQ vs Bayes-BT · IHQ-all | Top-1 accuracy | Improved |
| use very few raters | Crowd-BT · IHQ-all | convergence | Hurt |
Closest papers by meaning
- BayesJudge: Uncertainty-Aware Bayesian Meta-Evaluation of Human and LLM Judgments (NeurIPS 2026 · similarity 0.88)
- Bridging Human and LLM Judgments: Understanding and Narrowing the Gap (NeurIPS 2025 · similarity 0.86)
- From Uncertain Judgments to Calibrated Rankings: Conformal Elo Estimation for LLM Evaluation (NeurIPS 2026 · similarity 0.86)
- Rich Insights from Cheap Signals: Efficient Evaluations via Tensor Factorization (NeurIPS 2026 · similarity 0.86)
- A Unifying Information-theoretic Perspective on Evaluating Generative Models (AAAI 2025 · similarity 0.86)
- Maximizing Signal in Human-Model Preference Alignment (AAAI 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
