GUI Agent
WebArena: Why Does GPT-4 Achieve Only a 14.41% Success Rate on Realistic Web Tasks?
Categories
Categories
Explore posts organized by topic.
Harness Engineering
Self-Improving Agents Have an Overfitting Problem
Categories
Explore posts organized by topic.
WebArena: Why Does GPT-4 Achieve Only a 14.41% Success Rate on Realistic Web Tasks?
Self-Improving Agents Have an Overfitting Problem