AI Paper Atlas

Research map / Large language models / Alignment and preferences

ELSPR: Evaluator LLM Training Data Self-Purification on Non-Transitive Preferences via Tournament Graph Reconstruction

Yan Yu, Yilun Liu, Minggui He, Shimin Tao, Weibin Meng, Xinhua Yang, Li Zhang, Hongxia Ma, Dengye Li, Daimeng Wei, Boxing Chen, Fuliang Li

AAAI 2026 · Large language models › Alignment and preferences · approach: LLM-as-judge and feedback data · 2 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
fine-tune on cleaned (tournament-filtered) data instead of rawQwen2.5-7B-Instruct evaluator · cross-dataset averagepreference non-transitivity: 64.3% -> 50.5%Improved
fine-tune on cleaned data instead of rawQwen2.5-7B-Instruct evaluator · cross-dataset averagenormalized structural entropy: 0.811 -> 0.723Improved
fine-tune on cleaned data instead of rawQwen2.5-7B-Instruct evaluator · Helpful_Basepreference non-transitivity: 62.0% -> 44.9%Improved
fine-tune on cleaned data instead of rawLLaMA3.1-8B-Instruct evaluator · Helpful_Basepreference non-transitivity: 59.0% -> 40.2%Improved
fine-tune on targeted-filtered data instead of random 20% filtering (ablation)Qwen2.5-7B-Instruct evaluator · Helpful_Basepreference non-transitivity: 60.6% -> 44.9%Improved
fine-tune on random 20% filtered data instead of raw (ablation)Qwen2.5-7B-Instruct evaluator · Helpful_Basepreference non-transitivity: 62.0% -> 60.6%No clear effect
fine-tune on cleaned data instead of rawQwen2.5-7B-Instruct evaluator · Vicuna (tie-allowed prompt)preference non-transitivityMixed
fine-tune on cleaned data instead of rawQwen2.5-7B-Instruct evaluator · Vicuna (tie-allowed prompt)tie probability: 10.06% -> 6.13%Improved
keep cleaned instead of discarded preference pairspreference data · 5 datasetshuman consistency: 34.4% -> 52.6%Improved
keep cleaned instead of discarded preference pairspreference data · 5 datasetsmodel-human agreement: 51.2% -> 80.6%Improved
fine-tune on cleaned data instead of rawevaluator in MT-bench framework · Helpful_Basestd of adjusted win rates: 12.2% -> 16.1%Improved
fine-tune on cleaned data instead of rawevaluator in MT-bench framework · Self-instructstd of adjusted win rates: 8.2% -> 9.7%Improved

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