Benchmark RadarImaging-101
Imaging-101 evaluates coding agents on 57 computational imaging tasks across six scientific domains, with three tracks for planning, function-level unit tests, and end-to-end reconstruction.
- Released
- 2026-07-12
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
General coding benchmarks may not capture domain-specific challenges in scientific imaging, and this benchmark could help assess agent capabilities in algorithm selection, physical conventions, and pipeline integration.
Motivation
Computational imaging, which recovers hidden signals from indirect, noisy measurements, underpins quantitative discovery across scientific disciplines, yet building a correct reconstruction pipeline demands deep domain expertise and remains laborious even for domain scientists.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.