AutoSciBench: Autonomous Benchmark Generation for Evaluating Scientific Agents
AutoSciBench automatically generates scientific-agent benchmarks that lower solver accuracy versus human-curated tests.
1 stories, newest first · one card per story
AutoSciBench automatically generates scientific-agent benchmarks that lower solver accuracy versus human-curated tests.
Today 1.6K readers · 2.5K views · 779 reads · all time 189K views