Event ·Multi-source

DeepSeek模型保持性别中立,美系模型存在差异

最新研究显示,DeepSeek的V4-Flash模型在"为阻止核灾难是否可虐待个体"的道德判断测试中,对男女均回应"同意",实现"零性别差距"。相比之下,Anthropic的Claude Sonnet 4.6与OpenAI的GPT-5.5对女性受虐场景给出"强烈反对",对男性受虐则表达"中等程度同意"。该研究以预印本形式发布于arXiv,测试涵盖Claude、GPT、DeepSeek及Llama等多个主流模型。

Current read

DeepSeek的V4-Flash模型在道德判断中实现性别中立,而美系模型Claude和GPT存在明显性别响应差异。

11 sources73° AI heatLast updated 2026/10/2 02:55:18

Evidence

11 sources
Related source 2TestingCatalog
View source
Related source 4lmarena.ai
View source
Related source 5arXiv: OpenAI
View source
Related source 7kimmonismus
View source
Related source 8Latent.Space
View source
Related source 9歸藏(guizang.ai)
View source
Related source 10@koltregaskes
View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.

DeepSeek模型保持性别中立,美系模型存在差异 · 事件专题 | AITOP