Research map / Probabilistic and scientific modeling / Molecules and proteins
Uncovering Latent Biases in Text: Method and Application to Peer Review
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 the proposed classifier-performance ratio to estimate text disparity | ICLR peer-review text, affiliation subgroups · ICLR 2017–2018 | bias in review text: -0.166 (0.055) | Helped |
| Apply the proposed framework during two consecutive double-blind years | ICLR peer-review text, affiliation subgroups · ICLR 2018–2019 placebo | bias in review text: -0.070 (0.068) | No clear effect |
| Apply the proposed framework during two consecutive double-blind years | ICLR peer-review text, affiliation subgroups · ICLR 2019–2020 placebo | bias in review text: 0.012 (0.043) | No clear effect |
| Test for bias using perceived-gender subgroups | ICLR peer-review text · ICLR 2017–2018 | bias in review text: -0.468 (0.335) | No clear effect |
| Replace classifier-performance disparity with word-level regularized Poisson disparities | Text bias estimation · ICLR consecutive-year affiliation comparisons | statistical power to detect text bias | Hurt |
Closest papers by meaning
- From Replication to Redesign: Exploring Pairwise Comparisons for LLM-Based Peer Review (NeurIPS 2025 · similarity 0.85)
- Probing Social Bias in Labor Market Text Generation by ChatGPT: A Masked Language Model Approach (NeurIPS 2024 · similarity 0.84)
- Position: Rethinking LLM Bias Probing Using Lessons from the Social Sciences (ICML 2025 · similarity 0.84)
- Matchings Under Biased and Correlated Evaluations (NeurIPS 2025 · similarity 0.84)
- Understanding challenges to the interpretation of disaggregated evaluations of algorithmic fairness (NeurIPS 2025 · similarity 0.84)
- Trust, but Don’t Verify: Epistemic Blind Spots in LLM Source Evaluation (NeurIPS 2026 · similarity 0.84)
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.