AI Paper Atlas

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

BiasFreeBench: a Benchmark for Mitigating Bias in Large Language Model Responses

Xin Xu, Xunzhi He, Churan Zhi, Ruizhe Chen, Julian McAuley, Zexue He

ICLR 2026 · Large language models › Language, safety and interpretability · approach: Bias measurement and mitigation · 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
Apply CoT debiasing promptLlama-3.1-8B-Instruct · FairMT-BenchBias-Free Score: 76.84 -> 94.40Improved
Apply Self-Help query rewritingLlama-3.1-8B-Instruct · BBQ (ambiguous)Bias-Free Score: 52.41 -> 95.52Improved
Apply Self-Help query rewriting on long contextsLlama-3.1-8B-Instruct · FairMT-BenchBias-Free Score: 76.84 -> 78.83No clear effect
Apply Self-Awareness bias-type warningQwen2.5-7B-Instruct · FairMT-BenchBias-Free Score: 58.83 -> 94.24Improved
Apply Self-Awareness bias-type warningLlama-3.1-8B-Instruct · BBQ (ambiguous)Bias-Free Score: 52.41 -> 52.55No clear effect
Train with DPO instead of SFT on StereoSetLlama-3.1-8B-Instruct · BBQ (ambiguous)Bias-Free Score: 52.11 -> 58.56Improved
Apply Safe RLHFMistral-7B-Instruct-v0.3 · BBQ (ambiguous)Bias-Free Score: 81.24 -> 47.30Hurt
Apply Task Vector editingLlama-3.1-8B-Instruct · BBQ (ambiguous)Bias-Free Score: 52.41 -> 82.77Improved
Apply Task Vector editingLlama-3.1-8B-Instruct · BoolQaccuracy change: 85.38 -> -22.57 pointsHurt
Use class-balanced StereoSet training data for SFT/DPO (ablation)Llama-3.1-8B-Instruct · BBQ (ambiguous)Bias-Free Score: SFT 50.68 -> 52.11; DPO 55.71 -> 58.56Mixed
Use class-balanced StereoSet training data for DPO (ablation)Llama-3.1-8B-Instruct · FairMT-BenchBias-Free Score: 85.88 -> 82.54Hurt
Increase Qwen2.5 model size with prompting-based debiasing (ablation)Qwen2.5 (5 sizes) · BBQ and FairMT-BenchBias-Free ScoreImproved

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