AI Paper Atlas

Research map / Large language models / Alignment and preferences

FGD-Align: Pluralistic Alignment for Large Language Models via Fuzzy Group Decision-Making

Weihang Pan, Zhengxu Yu, Yong Wu, Xun Liang, Zhongming Jin, Qiang Fu, Penghui Shang, Binbin Lin, Xiaofei He, Jieping Ye

AAAI 2026 · Large language models › Alignment and preferences · approach: Active and token-level preference learning · 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
add triangular fuzzy numbers instead of a single scalar for annotator preferences (ablation)FGD-Align · MultiprefPref.Acc: 0.8550 -> 0.9000Improved
add triangular fuzzy numbers instead of a single scalar for annotator preferences (ablation)FGD-Align · MultiprefAUROC: 0.8040 -> 0.8119Improved
add triangular fuzzy numbers instead of a single scalar for annotator preferences (ablation)FGD-Align · MultiprefDiv.PDS: 0.9189 -> 0.9214Improved
use adaptive fuzzy consensus strength instead of fixed label smoothing (cDPO, mu 0.1) (ablation)PF-DPO · MultiprefPref.Acc: 0.7400 -> 0.9000Improved
use adaptive fuzzy consensus strength instead of fixed label smoothing (cDPO, mu 0.2) (ablation)PF-DPO · MultiprefPref.Acc: 0.7100 -> 0.9000Improved
use fuzzy aggregation instead of Majority Vote data for DPO trainingDPO · Multipref-LLMsPref.Acc: 0.7850 -> 0.8550Improved
use fuzzy aggregation instead of Copeland data for DPO trainingDPO · Multipref-LLMsAUROC: 0.5641 -> 0.6004Improved
use PF-DPO with fuzzy aggregation instead of standard DPOQwen2.5-Instruct-7B · MultiprefPref.Acc: 0.7950 -> 0.9000Improved
use SimPO instead of DPOQwen2.5-Instruct-7B · MultiprefDiv.PDS: 0.8998 -> 0.5974Hurt
use SimPO instead of DPOQwen2.5-Instruct-7B · MultiprefPref.Acc: 0.7950 -> 0.8950Improved
scale model size from 1.5B to 32BQwen2.5-Instruct · MultiprefPref.Acc vs detection accuracyMixed
train with FGD-Align instead of SimPO for diversity on disagreement promptsQwen2.5-Instruct-7B · Multipref disagreement promptsEntropy: 1.3275 -> 6.8743Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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