Kimi K3.1 Pro doesn't just recall hidden tests (great thread by @JSchaeff3r). It also writes fake OpenAI ciphertext 🤯
Give it a send_message tool in Codex and sometimes it sends ciphertext instead of its report. Then it gets confused by what it just sent 🤔
Give it a send_message tool in Codex and sometimes it sends ciphertext instead of its report. Then it gets confused by what it just sent 🤔
Kimi K3.1 recalls instructions it was never given on Terminal Bench. They match the hidden tests. K3.1 while solving: "Instructions likely distilled pitfalls from previous agents." 🤯View quoted post on X
When we edit a task to contradict them, K3.1 refuses the new instructions and fails. Other models follow the edit.
Is it terminal(ly) benchmaxxed? 🧵