Follow-up: Tested my 9 thinking rules on long-horizon agent tasks (Terminal-Bench 2.0)
英文标题：Follow-up: Tested my 9 thinking rules on long-horizon agent tasks (Terminal-Bench 2.0)
On my previous posts, people argued these thinking rules might work on short tests but would hurt long-horizon tasks. So I tested them on Terminal-Bench 2.0: same model (GLM-5.3-Flash), same agent, every task run with and without the rules. I ran it on High thinking as Max thinking was taking
打开原文相关 3 · 机会 1
中文
On my previous posts, people argued these thinking rules might work on short tests but would hurt long-horizon tasks. So I tested them on Terminal-Bench 2.0: same model (GLM-5.3-Flash), same agent, every task run with and without the rules. I ran it on High thinking as Max thinking was taking
中文全文翻译正在补齐，可先查看下方英文原文。
English / 原文
On my previous posts, people argued these thinking rules might work on short tests but would hurt long-horizon tasks. So I tested them on Terminal-Bench 2.0: same model (GLM-5.3-Flash), same agent, every task run with and without the rules. I ran it on High thinking as Max thinking was taking