arXiv AI

One Example Is Enough to Pass Fairness Benchmarks: Rethinking Fairness Evaluation for Aligned LLMs

The paper demonstrates that a single example from the BBQ fairness benchmark can dramatically improve a model’s performance, raising accuracy from 79.9% to 92.9% with Group Relative Policy Optimization and to 99.0% with one-shot in-context learning. This effect is consistent across different model families and is driven by the model’s reasoning traces, which adopt a category‑agnostic "missing evidence" pattern. The authors argue that BBQ-style multiple‑choice abstention tests capture only a single structural cue and therefore do not guarantee true fairness, calling for broader evaluation suites.

arXiv Machine Learning

Aug 10

Let's Unlearn Stereotypes Before Decision-Making: Assessing the Impact of Intrinsic Bias Mitigation on Downstream Fairness in LLMs

arXiv:2509. 16462v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are increasingly used in high-stakes decision-making systems, where biased predictions can reinforce social and economic disparities.

By Mina Arzaghi, Alireza Dehghanpour Farashah, Florian Carichon, Jean-Fran\c{c}ois Plante, Golnoosh Farnadi

llms rag fine-tuning benchmarks safety

arXiv AI

Jul 24

Unlearning Under Imbalance: Benchmarking Fairness in Multimodal LLM Unlearning

arXiv:2607. 21300v1 Announce Type: cross Abstract: Machine unlearning has emerged as a tool for removing personal data from trained models to comply with recent AI regulations.

By Lorenzo Orsingher, Thomas De Min, Massimiliano Mancini, Davide Talon, Elisa Ricci

llms nlp fine-tuning multimodal benchmarks safety

arXiv AI

Jun 10

Is Fairness Truly Fair? Towards Reliable Lipschitz Fairness in Multi-Task Learning via Fixed-\texorpdfstring{$\delta$}{delta} Alignment

arXiv:2606. 10632v1 Announce Type: cross Abstract: Lipschitz-style individual fairness formalizes the idea that semantically similar examples should receive similar predictions, but its evaluation in multi-task learning (MTL) can be confounded by method-induced representation scales.

By Junbo Ding, Xin Zang, Chenchen Pan, Donghao Song, Jiaxin Zhu, Danhuai Guo

benchmarks safety

arXiv AI

Aug 17

Training Fair Tabular Foundation Models

arXiv:2608. 14211v1 Announce Type: cross Abstract: Tabular Foundation Models (TFMs) have emerged as leading methods for tabular predictive tasks, leveraging in-context learning to predict on new data without task-specific training.

By Patrik Kenfack, Jesse C. Cresswell, Anthony L. Caterini, Samira Ebrahimi Kahou, Ulrich A\"ivodji

llms safety

arXiv AI

Aug 3

Benchmarks Are Not Monolithic: Sample-Level Auditing and Orchestration for LLM Evaluation

arXiv:2607. 28801v1 Announce Type: cross Abstract: Benchmark datasets are central to evaluating Large Language Models (LLMs), yet they are typically conceived as monolithic tasks, obscuring substantial variation in the demands of individual samples.

By Philipp D. Siedler, Jordan Sassoon

llms agents benchmarks safety

arXiv AI

Jun 11

Soft-Prompt Tuning for Fair and Efficient LLM Benchmark Evaluation

arXiv:2606. 12117v1 Announce Type: cross Abstract: Benchmark scores often misrepresent a large language model's (LLM's) knowledge, because they rely, e.

By Selen Erkan, Bastian Boll, Kristian Kersting, Bj\"orn Deiseroth, Letitia Parcalabescu

llms benchmarks

arXiv AI

Jul 7

Best-of-Better-$N$: Generating Pre-Aligned Responses with In-Context Learning

arXiv:2607. 03453v1 Announce Type: cross Abstract: Inference-time alignment methods, such as Best-of-$N$, offer a flexible alternative to training-based alignment by using reward models to select high-quality responses generated by a reference LLM.

By Eric Lei, Hsiang Hsu, Chun-Fu Chen

llms benchmarks safety

arXiv AI

Jul 22

RRPO: Reference-Relative Policy Optimization with Stratified Conditional Rollouts

arXiv:2607. 18470v1 Announce Type: cross Abstract: Group Relative Policy Optimization (GRPO) has shown strong effectiveness in reinforcement learning from verifiable feedback, where sampled rollouts can be compared within a group using task-provided correctness signals.

By Yuxin Xiong, Xunyi Jiang, Rohan Surana, Xintong Li, Sheldon Yu, Nikki Lijing Kuang, Ryan A. Rossi, Jingbo Shang, Tong Yu, Julian McAuley, Junda Wu

reinforcement-learning fine-tuning safety

arXiv Computation and Language

Sep 15

Intersectional Fairness in Large Language Models

arXiv:2604.20677v3 Announce Type: replace Abstract: Large Language Models (LLMs) are increasingly deployed in socially sensitive settings, raising concerns about fairness and bias, particularly when...

By Chaima Boufaied, Ronnie De Souza Santos, Ann Barcomb

llms nlp benchmarks safety

arXiv AI

Sep 16

FairLint-DL: An IDE-Native Tool for Fairness Debugging of Deep Learning Software

arXiv:2609.16321v1 Announce Type: cross Abstract: Existing fairness analysis tools predominantly operate as post-training evaluation frameworks, requiring practitioners to complete the full model dev...

By Archit Rathod, Saeid Tizpaz-Niari

benchmarks safety

arXiv AI

Jun 4

BiasGRPO: Stabilizing Bias Mitigation in High-Variance Reward Landscapes via Group-Relative Policy Optimization

arXiv:2606. 04807v1 Announce Type: new Abstract: Mitigating social bias in Large Language Models (LLMs) presents a distinct alignment challenge: unlike verifiable tasks, bias lacks a single ground truth, creating a high-variance, subjective reward landscape.

By Saket Reddy, Ke Yang, ChengXiang Zhai

llms reinforcement-learning fine-tuning benchmarks safety

arXiv AI

Jun 2

Demystifying the Optimal Fair Classifier in Multi-Class Classification

arXiv:2606. 00656v1 Announce Type: cross Abstract: Ensuring fair and equitable treatment across diverse groups, particularly in multi-class classification tasks, poses a significant challenge due to the persistent biases inherent in machine learning models.

By Li Zhang, Yuyuan Li, XiaoHua Feng, Jiaming Zhang, Fengyuan Yu, Chaochao Chen

fine-tuning safety

Related stories