AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Format as a Prior: Quantifying and Analyzing Bias in LLMs for Heterogeneous Data

Jiacheng Liu, Mayi Xu, Qiankun Pi, Wenli Li, Ming Zhong, Yuanyuan Zhu, Mengchi Liu, Tieyun Qian

AAAI 2026 · Large language models › Language, safety and interpretability · approach: Uncertainty quantification for LLMs · 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
convert homogeneous inputs into heterogeneous formats with equivalent contentLLM QA · HotpotQAaccuracy: -9%Hurt
convert homogeneous inputs into heterogeneous formats with equivalent contentLLM QA · MuSiQueaccuracy: -12%Hurt
present both conflicting evidences in same format (plain text) instead of different formats (ablation)ten LLMs · ConflictBank pairsDCR: heterogeneous 3.01%-24.02%, markedly higher when homogeneousImproved
increase model size within Qwen3 series (ablation)Qwen3 · ConflictBank pairsDCRNo clear effect
increase entries in structured evidence (4 -> 8 -> 12) vs texts (ablation)KG evidence · Information richness, KGs vs textsFPR for structured: 0.334 -> 0.412 -> 0.472Improved
increase entries in structured evidence (4 -> 12) vs texts (ablation)Table evidence · Information richness, tables vs textsFPR for structured: 0.238 -> 0.366Improved
corrupt structure tokens of structured evidence (ablation)structured evidence vs texts · Structure qualityFPR for structuredHurt
add attention-balancing reweighting at decodingQwen3-8B, Mistral-7B, Llama-3.1-8B · ConflictBank pairsDCRImproved
add attention-balancing reweighting at decodingQwen3-8B, Mistral-7B, Llama-3.1-8B · ConflictBank pairsFPRNo clear effect
add attention-balancing reweightingheterogeneous RAG-style QA · HotpotQAaccuracy: +6.5%Improved
add attention-balancing reweightingheterogeneous RAG-style QA · MuSiQueaccuracy: +9.5%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