Benchmark RadarAI BENCHMARK PROFILE
VisTarget-Bench
A 150-task human-verified benchmark pairing questions with held-out target images to separate image-retrieval failures from visual-perception failures in multimodal search agents.
- Released
- 2026-08-28
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
Targets a gap in evaluating visual grounding within search agent trajectories, enabling finer diagnosis of retrieval versus perception errors.
Motivation
Multimodal search agents extend parametric knowledge with newly emerging and long-tail evidence from the open web.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.