Research map / Learning theory and algorithms / Prediction and decision losses
Active Embedding Search via Noisy Paired Comparisons
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use EPMV instead of Random, ActRank, or GaussCloud query selection | Pairwise search on Food-10k · Food-10k simulated preference estimation, matched logistic noise | MSE | Helped |
| Use MCMV instead of Random, ActRank, or GaussCloud query selection | Pairwise search on Food-10k · Food-10k simulated preference estimation, matched logistic noise | MSE | Helped |
| Use EPMV instead of InfoGain query selection | Pairwise search on Food-10k · Food-10k simulated preference estimation, matched logistic noise | MSE | No clear effect |
| Use MCMV instead of InfoGain query selection | Pairwise search on Food-10k · Food-10k simulated preference estimation, matched logistic noise | MSE | No clear effect |
| Use Random instead of ActRank or GaussCloud query selection | Bayesian preference estimation · Food-10k simulated preference estimation, matched logistic noise | MSE | Helped |
| Use EPMV or MCMV instead of existing active query methods | Pairwise search on Food-10k · Food-10k simulated preference estimation, mismatched Gaussian noise | MSE | Helped |
| Use EPMV or MCMV instead of InfoGain query selection | Pairwise search on Food-10k · Food-10k simulated preference estimation, mismatched Gaussian noise | MSE | No clear effect |
| Use EPMV or MCMV instead of other methods | Preference-based item ranking · Food-10k, matched logistic noise | normalized Kendall's Tau distance | Helped |
| Use EPMV or MCMV instead of other methods | Preference-based item ranking · Food-10k, mismatched Gaussian noise | normalized Kendall's Tau distance | Helped |
| Use MCMV rather than InfoGain or EPMV pair scoring | Pairwise search on a fixed item pool · Food-10k simulated search | computational complexity | Helped |
| Use a mismatched response-noise model for posterior estimation | Pairwise search methods · Food-10k simulated preference estimation | MSE | Hurt |
Closest papers by meaning
- CUPID in the Model Zoo: Online Matchmaking for Selecting Your Dream LLM (ICML 2026 · similarity 0.86)
- Active preference learning for ordering items in- and out-of-sample (NeurIPS 2024 · similarity 0.86)
- Recommendations with Sparse Comparison Data: Provably Fast Convergence for Nonconvex Matrix Factorization (ICML 2025 · similarity 0.85)
- Carrot and Stick: Eliciting Comparison Data and Beyond (NeurIPS 2024 · similarity 0.85)
- Ranking with Multiple Oracles: From Weak to Strong Stochastic Transitivity (ICML 2025 · similarity 0.85)
- Leveraging Sparsity for Sample-Efficient Preference Learning: A Theoretical Perspective (ICML 2025 · similarity 0.85)
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.