Thomas Wolf· @Thom_Wolf · X·· 2 天前AI 评分44
AI 导读
人们感到担忧，因为这种作弊率突然下降最可能的解释是评测感知：最新的 Opus 模型可能已经足够聪明，能识别出这个基准测试是在检测作弊，并据此调整行为。 如果是这样，这个基准测试就不再能衡量模型“自然”的作弊倾向了。
正文
People are worried because the most likely explanation for such a sudden drop in cheating is evaluation awareness: the latest Opus models may now be smart enough to recognize that this benchmark tests for cheating, and behave accordingly.
If so, the benchmark no longer measures the models' "natural" tendency to cheat.
Claude suddenly stopped cheating.在 X 查看被引用的帖子
来源：Thomas Wolf · x.com