Story
arxiv_cs_cl · Sep 16, 2026 · paper
Source brief
ReFigBench: Benchmarking Scientific Figure Reconstruction as Editable PowerPoint Artifacts
arxiv.orgSep 16, 2026
original source linked
In brief
Multimodal coding agents are expected to turn visual inputs into usable artifacts, and they act through a harness, the layer of tools, context management, and execution environment around the model. Existing evaluatio...
Feed lens
agentharnessevaluation