跳到正文
原文
The Decoder:AI News· Matthias Bastian·· 9 天前AI 评分31

Thinking Machines 的 Inkling 2 Small 在基准测试思考中出现 316 处脏话

Thinking Machines' Inkling 2 Small drops 316 F-bombs across benchmark runs

文章要点
  • 1研究方 Vals AI 检查 13 项基准共 2,655 次运行的思考日志,发现脏话 316 处、涉及 50 次运行
  • 2241 处集中在两次失控运行,一次小行星轨迹计算生成约 50 万字符回复,重复句子并含 158 处脏话
  • 3脏话不能当失败预警,96% 的失败运行没有脏话,全部计分运行本身失败率约 75%
  • 4模型多在自责、自以为突破或放弃时骂脏话,用户通常看不到思考内容或只能过滤查看
决策要点

脏话是模型失败时的行为痕迹而非可靠信号,思考过程文本不适合作为任务监控的依据。

  • 影响:靠读模型思考内容来发现任务失败的设想不成立,质量监控仍要以输出结果为准;若产品把思考原文展示给终端用户,此类内容会构成体验和品牌风险。
  • 参考:可考虑为智能体任务配置基于结果校验的监控,不依赖过程文本告警;值得排查自家产品是否会把思考原文暴露给用户,必要时加过滤。
  • 依据:数据来自 Vals AI 对 2,655 次基准运行思考日志的第三方分析,暂无独立验证;约 75% 失败率为其基准场景数字。
正文

Thinking Machines' Inkling 2 Small swears in its internal chain of thought. Researchers at Vals AI read the model's thinking logs from 2,655 runs across 13 benchmarks and found "fuck" or variations in 50 runs, totaling 316 instances. 241 came from just two runs that spiraled out of control. In one, tasked with calculating an asteroid's trajectory, the model produced a single 500,000-character response, roughly novel-length, repeating the same sentences with 158 expletives. ProgramBench was hit hardest, with 20 of 200 runs affected.

Who hasn't felt like this?  | Image: via ValsAI

Inkling 2 Small typically swears about its own mistakes, apparent breakthroughs, or when it gives up. One run read: "FUCK. SUBMIT. DONE. IT'S NOT WORKING. NOTHING WORKS." The profanity doesn't work as a warning signal, though. While 46 of 48 scored runs with swearing failed, about 75 percent of all scored runs fail anyway, and 96 percent of failed runs contained no expletives. Users typically don't see these thinking steps, or only with a filter.

AI News Without the Hype – Curated by Humans

Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section.

来源:The Decoder:AI News · the-decoder.com