Anthropic @AnthropicAI
3) Finding cases of inverse scaling in test-time compute – where more and more reasoning leads to worse and worse outcomes:
https://t.co/ZgnK1wECw5
https://t.co/ZgnK1wECw5
Aryo Pradipta Gema @aryopg
New Anthropic Research: “Inverse Scaling in Test-Time Compute”
We found cases where longer reasoning leads to lower accuracy.
Our findings suggest that naïve scaling of test-time compute may inadvertently reinforce problematic reasoning patterns.
🧵
We found cases where longer reasoning leads to lower accuracy.
Our findings suggest that naïve scaling of test-time compute may inadvertently reinforce problematic reasoning patterns.
🧵
