2026_09
We released RSI-Exam 1.0, a comprehensive benchmark that evaluates whether an AI agent can improve itself over a long horizon and generalize to unseen data. Details at: https://rsi-exam.ai/.
We released RSI-Exam 1.0, a comprehensive benchmark that evaluates whether an AI agent can improve itself over a long horizon and generalize to unseen data. Details at: https://rsi-exam.ai/.