We introduce PaperBench, a benchmark evaluating the ability of AI agents to replicate state-of-the-art AI research.
Related Posts
Chatham scales its capital markets expertise with OpenAI
Chatham Financial uses Codex and GPT-5.6 to build technology and redesign workflows, cutting trade validation from 30 minutes to under…
Building a safer path to autonomous industrial AI
Industrial AI is entering a new phase. After decades of predictive analytics and other specialized applications, advances in foundation models,…
Applications of AI at OpenAI
Explore how OpenAI products like ChatGPT, Codex, and APIs bring AI into real-world use for work, development, and everyday tasks.