1 articles with this tag
Rishi Desai from Abundant AI introduces SWE-Marathon, a benchmark evaluating AI coding agents on billion-token scale tasks, revealing current limitations and the need for robust verification.
