arXiv Computation and Language

Alignment Reduces Expressed but Not Encoded Gender Bias: A Unified Framework and Study

The paper introduces a unified framework that simultaneously measures intrinsic (encoded) and extrinsic (expressed) gender bias in large language models using identical neutral prompts. It finds a consistent link between latent gender information and output bias, but shows that alignment via supervised fine‑tuning reduces expressed bias while leaving internal gender associations largely intact and reactivatable by adversarial prompts. The study also demonstrates that debiasing gains on structured benchmarks may not transfer to realistic tasks such as story generation.

arXiv Computer Vision

Sep 16

ViD: Vision-Dominant Gender Bias Mitigation for Large Vision-Language Models

The paper introduces ViD, a vision‑dominant gender bias mitigation framework for large vision‑language models. ViD uses causal analysis of attention patterns and dual mechanisms—backdoor adjustment and refined token selection—to suppress bias while preserving reasoning and generation quality. Experiments show a 14.7% reduction in gender bias on FACET and significant improvements on MS COCO image captioning, all without extra training overhead.

By Zhipeng Zhao, Zhaoqiang Wei, Peishun Liu, Youwei Zhao, Ruichun Tang

llms multimodal benchmarks safety

arXiv AI

Jun 2

Failure of contextual invariance in large language models

arXiv:2603. 23485v2 Announce Type: replace-cross Abstract: Standard evaluation practices assume that large language model (LLM) outputs are stable when prompts are embedded in contextually equivalent discourses.

By Sagar Kumar, Ariel Flint, Luca Maria Aiello, Andrea Baronchelli

llms benchmarks safety

arXiv Computation and Language

Oct 19

Analyzing Occupational Gender Bias across the LLM Training Lifecycle

arXiv:2610.23587v1 Announce Type: new Abstract: Post-training is widely assumed to reduce the social biases language models inherit from pretraining, yet direct evidence for this assumption remains m...

By Manthan Gadhia, Alexander Hoyle, Karolina Sta\'nczak

llms reinforcement-learning fine-tuning safety

arXiv AI

Oct 15

From Individuals to Interactions: Benchmarking Gender Bias in Multimodal Large Language Models from the Lens of Social Relationship

arXiv:2506.23101v2 Announce Type: replace-cross Abstract: Ensuring the trustworthiness of Multimodal Large Language Models (MLLMs) requires rigorous evaluation of gender bias, particularly in sociall...

By Yue Xu, Wenjie Wang

llms multimodal benchmarks safety

arXiv Machine Learning

Jul 23

Evaluating and Mitigating Gender Bias in Pre-trained Embeddings for ML-based Recruitment

arXiv:2607. 20073v1 Announce Type: new Abstract: AI-based recruitment systems that rely on machine learning models trained on historical CV data, risk perpetuating and amplifying social biases.

By Farnaz Faramarzi Lighvan, Lynn Houthuys

llms rag safety

arXiv Computer Vision

Sep 24

Gender Bias in Vision-Language In-Context Learning

The paper investigates how in‑context learning (ICL) in large vision‑language models (LVLMs) can amplify gender bias. Using the VL‑BICLE framework, the authors show that gendered ICL demonstrations shift model bias toward the demonstrated gender, especially in tasks involving gendered language such as image captioning and pronoun prediction. They find that similarity‑based retrieval does not mitigate this bias and that replacing real images with synthetic ones from stable diffusion reduces bias without hurting caption quality.

By Tong Xiang, Noa Garcia, Yuta Nakashima

llms diffusion nlp multimodal safety

arXiv AI

Jul 8

RPAM: A Principled Metric for Evaluating Associations in Language Models with High Predictive Validity in Downstream Outputs

arXiv:2607. 05679v1 Announce Type: cross Abstract: Language models (LMs) exhibit problematic biases, such as stereotypes.

By Damian Hodel, Jevin West, Aylin Caliskan

llms rag benchmarks

arXiv AI

Aug 18

DR.GAP: Mitigating Bias in Large Language Models using Gender-Aware Prompting with Decoupled Reasoning

arXiv:2502. 11603v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) exhibit strong natural language understanding capabilities but also inherit and amplify societal biases, particularly gender bias, raising fairness concerns.

By Hongye Qiu, Yue Xu, Yi Wang, Meikang Qiu, Wenjie Wang

llms multimodal safety

arXiv Computation and Language

Sep 21

Benchmarking Gender Bias in Machine Translation Evaluation Metrics across Occupations

The paper investigates gender bias in machine translation evaluation metrics using an occupation-balanced subset of GAMBIT+ across seven English‑source language pairs, including a new German extension. It finds that masculine translations tend to receive higher scores and that biases align with stereotypical gender representations, though the strength varies by evaluator and language. The study highlights that assessing bias requires multiple dimensions beyond a single aggregate measure.

By Orfeas Menis Mastromichalakis, Giorgos Filandrianos, Wafaa Mohammed, Giuseppe Attanasio, Chrysoula Zerva

nlp benchmarks safety

arXiv AI

Sep 1

Man Made Language Models? Evaluating LLMs' Perpetuation of Masculine Generics Bias

arXiv:2502.10577v2 Announce Type: replace-cross Abstract: Instruct-based large language models (LLMs) have been shown to propagate and even amplify gender bias when prompted with contextually constra...

By Enzo Doyen, Amalia Todirascu

llms safety

arXiv AI

Sep 2

BiasGym: A Simple and Generalizable Framework for Analyzing and Removing Biases through Injection

BiasGym is a cost‑effective, generalizable framework that injects specific biases into large language models via token‑based fine‑tuning while keeping the model frozen. It then uses two debiasing methods—Scope and Steer—to identify and suppress or redirect the components responsible for biased behavior. The framework enables consistent bias elicitation, precise localization of bias associations, and targeted debiasing without harming downstream performance, and it has been shown to reduce real‑world stereotypes such as labeling Italians as reckless drivers.

By Sekh Mainul Islam, Nadav Borenstein, Siddhesh Milind Pawar, Haeun Yu, Arnav Arora, Isabelle Augenstein

llms fine-tuning safety

arXiv AI

Sep 1

Vision-Language Models Suppress Female Representations Under Ambiguous Input

arXiv:2605.31556v2 Announce Type: replace-cross Abstract: Alignment teaches vision-language models (VLMs) to avoid expressing demographic biases, and when gender is clearly visible they largely succe...

By Arnau Marin-Llobet, Simon Henniger, Mahzarin R. Banaji

llms rag multimodal benchmarks safety

Related stories