AI Paper Atlas

Research map / Large language models / Alignment and preferences

UDA: Unsupervised Debiasing Alignment for Pair-wise LLM-as-a-Judge

Yang Zhang, Cunxiang Wang, Lindong Wu, Wenbo Yu, Yidong Wang, Guangsheng Bao, Jie Tang

AAAI 2026 · Large language models › Alignment and preferences · approach: Robust and calibrated DPO · 1 citations (OpenAlex)

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 changedWhereMeasuredEffect
replace fixed K Elo with UDA adaptive K and soft labelsElo judging · ArenaHard, 10 judgesavg inter-judge std dev: 158.5 -> 64.8Improved
apply UDA adapter zero-shot to unseen promptsElo judging · Human-Annotated Transfer Setavg inter-judge std dev: 193.95 -> 70.97Improved
apply UDAElo judging · Human-Annotated Transfer SetPearson correlation with human: 0.651 -> 0.812Improved
add judge self-awareness features (judge's own answer embedding) (ablation)UDA · Human-Annotated Transfer SetPearson correlation with human: 0.510 -> 0.812Improved
add judge self-awareness features (ablation)UDA · Human-Annotated Transfer Setavg inter-judge std dev: 65.64 -> 70.97Hurt
apply UDA to reduce self-preferenceElo judging · curated dataset, self-score deviationself-score deviation from other judges: -21% to +56% -> -15% to +4%Improved
pull judges toward consensusUDA · well-calibrated judgesindividual accuracyHurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy