gemini-4IMAGE
幻觉率榜单：Gemini 4 Argon 15% 最低
一张对比图称 Gemini 4 Argon 幻觉率 15%，低于 GPT-4.1 Nano 的 18%，而 GPT-6 Fable 与 Claude Fable 5.1 高达 92%–95%。
KRISHNA
@KrishnaBuildsAI19 天前收录原推发布于 2026.10.17
KRISHNA @KrishnaBuildsAI原推发布于 2026.10.17
Wait… everyone talks about context windows and output tokens. But hallucination rates might be the benchmark that actually matters. Gemini 4 Argon: 15% GPT-4.1 Nano: 18% DeepSeek R2: 21% Llama 4 Scout: 24% GPT-6 Fable: 92% Claude Fable 5.1: 95% Lower is better. Not context. Not token limits. Accuracy. Why isn’t everyone talking about this?
x.com/krishnabuildsai/status/2111373968070611240采集于 2026.10.18 10:05