Ofir Press(@OfirPress):https://t.co/PAIAk8klpg 1. Use tasks that would take teams of humans years or decades to solve 2. Find real-world tasks that real-humans had to actually solve, don't just make stuff up or have humans make stuff up for you 3. Find tasks that can be scalably crawled from the web

Ofir Press

@OfirPress

SWE-bench, ProgramBench, SWE-sweep. Postdoc @Princeton. PhD @nlpnoah @UW.

Joined June 2016

10K Following 20.4K Followers

Ofir Press@OfirPress

2026.10.13 14:41

1. Use tasks that would take teams of humans years or decades to solve 2. Find real-world tasks that real-humans had to actually solve, don't just make stuff up or have humans make stuff up for you 3. Find tasks that can be scalably crawled from the web