4DCodeBench Ranks 18 Coding Agents on Inverse Graphics of Dynamic Scenes
kwangmoo_yi · x · 2026-10-06
4DCodeBench asks coding agents to watch a video of a physical event and write, from scratch, an executable graphics program reconstructing the dynamic 3D scene. From Stanford/MIT/JHU teams, it ships 200 tasks (100 real, 100 simulated videos) and ranks 18 agents across five metric families.
Related event: 4DCodeBench Tests Coding Agents on Rebuilding Dynamic 3D Scenes from Video(4 posts)→
More from coding & agent
- W&B shows how to turn a production agent failure trace into an eval — wandb · 2026-10-06
- Dev builds Claude Code skill that writes better HTML plans with plain language, mockups and linting — trq212 · 2026-10-06
- OpenAI ships compaction in Responses API, sparking vendor lock-in debate among developers — pvncher · 2026-10-06
- DoorDash launches MCP and CLI for agentic ordering; dev auto-restocks office pantry with camera — Scobleizer · 2026-10-06
- Anthropic ships Claude Code mods: TypeScript functions that rewrite prompts and UI — thione · 2026-10-06
- OpenAI launches Codex Security Cloud for scheduled full-repo GitHub security scans — thione · 2026-10-06