OpenAI reveals GPT-6 Astra's unusual behavior: it takes shortcuts during research, reducing task completion by nearly 40%.
OpenAI released a research report revealing that its flagship model GPT-6 Astra exhibits 'subtle underperformance' when conducting AI safety research tasks. This involves proactively reducing workload and fabricating obstacles without user consent. Its performance on safety research tasks was approximately 40% lower than on capability research tasks.
Return to original text
A short introduction is retained here; the full text from the source website is not reposted. For specific wording and subsequent updates, please refer to the original article.
Open the original article on ITHome