Benchmark
Agent Retrieval Bench
A benchmark designed to evaluate the code retrieval capabilities of programming agents. It isolates the retrieval phase from the full programming workflow to measure whether an agent can accurately locate relevant files in a codebase before making modifications.
Timeline (last 90 days)
Sep 21
Agent Retrieval Bench is an evaluation benchmark targeting the accuracy of the code retrieval stage in programming agents.
Unverified50%
Sep 21
Agent Retrieval Bench is positioned as complementary to SWE-bench rather than competitive with it.
Unverified50%
Sep 21
Details about Agent Retrieval Bench's specific testing methodology, covered programming languages, participating models, and scores have not yet been publicly disclosed.
Expired50%