WorldBox Boosts Agent Maze Exploration to 230/256 Rooms, 2x the GPT-6 Astra Baseline
ZhitingHu · x · 2026-10-16
EnactraAI showed WorldBox's impact on a maze exploration benchmark. Rooms visited by run's end: Opus 5.5 at 68, GPT-6 Astra at 113, and GPT-6 Astra + WorldBox at 230 of 256 (90%), stepping on nearly twice as many distinct tiles as the Astra baseline.
How WorldBox works: the agent starts with empty memory and no skills, writes its own notes and Python skills, reconstructs room geometry from camera views, and tests candidate moves in a sandbox — remembering where it's been sharply improves exploration decisions.
Related event: WorldBox Sandbox Doubles GPT-6 Astra's MazeBench Score to 61(3 posts)→
More from coding & agent
- Every publishes 8,000-word Codex oral history, agent-native so your AI can dig the raw interviews — danshipper · 2026-10-16
- Developer rebuilds GitHub in 4 days: Keel makes agents declare intent before changing code — threepointone · 2026-10-16
- Stop using small models as judges with overloaded context, dev warns — xeophon · 2026-10-16
- Meta's open-weight Muse Spark Open hits Ollama with 1M-token context and Claude Code support — ollama · 2026-10-16
- Monid launches: one balance, pay-per-call access to 1,700+ tools for AI agents — HeyAmit_ · 2026-10-16
- OpenAI publishes Sites tutorial: build and share your first Site — OpenAIDevs · 2026-10-16