HERA co-evolves agent harnesses to abstain on infeasible tasks
An evolved harness raised abstention accuracy from 61.7% to 83.3% and transferred to 19 other LLMs.
1 stories, newest first · one card per story
An evolved harness raised abstention accuracy from 61.7% to 83.3% and transferred to 19 other LLMs.
Today 1.6K readers · 2.5K views · 778 reads · all time 189K views