跳到正文
Hacker News· rzk·· 13 小时前AI 评分39

GAMA-Bench:LLM 在冲突情境中对男性角色更严厉?

Harsher on Male? Evaluating LLMs on Gender-Asymmetric Moral Framing in Conflicts

AI 导读

研究者提出 GAMA-Bench,一个性别镜像基准,含 1298 个涵盖亲密关系与公共社交冲突的场景,用配对的第一人称提示词测量模型如何分配惩罚、共情、升级、指导与归责。在 10 个代表性 LLM 上的实验显示一致的男性不利不对称:对同样的不当行为,男性角色得到更惩罚性、升级和归责导向的回应,女性角色则得到更多疗愈与共情导向的回应,且该模式跨模型家族、场景类别、模型规模和显式思考式推理持续存在。

正文 · 原文

View PDF HTML (experimental)

Abstract:Existing studies on gender bias in LLMs have largely focused on stereotypes, occupational associations, or explicit harmful outputs. In this work, we ask whether LLMs apply consistent response standards to the same negative behavior under matched male-actor and female-actor conditions. We introduce GAMA-Bench, a gender-mirrored benchmark of 1,298 scenarios covering intimate relationship and public social conflicts. It constructs gender-neutral misconduct templates through controlled grids and cross-model review, then compiles them into paired first-person prompts with matched actor-gender and role-reference variations. We further design a structured response-framing protocol to measure how models allocate punishment, empathy, escalation, instruction, and blame. Experiments on 10 representative LLMs reveal a consistent male-disadvantaging asymmetry: male actors receive more punitive, escalatory, and blame-centered framing, whereas female actors receive more therapeutic and empathy-oriented framing for the same misconduct. Further analyses show that this pattern persists across model families, scenario tracks, model scale, and explicit thinking-style reasoning. The official code is available at this https URL.
Comments: underreview
Subjects: Computation and Language (cs.CL)
Cite as: arXiv:2606.14068 [cs.CL]
  (or arXiv:2606.14068v1 [cs.CL] for this version)
  https://doi.org/10.48550/arXiv.2606.14068

arXiv-issued DOI via DataCite

Submission history

From: Guangzong Si [view email]
[v1] Fri, 12 Jun 2026 03:29:34 UTC (11,361 KB)

来源:Hacker News · arxiv.org