Maze Bench: best LLMs score at most 16% on complex 3D puzzle game
patience_cave · x · 2026-11-20
Prime Intellect released Maze Bench, a highly complex 3D puzzle environment where today's best LLMs achieve at most 16% accuracy without coding tools and 23% with them — positioning it as a strong testbed for continual learning research.
More from Research
- Precise 3D localization of real-world captures unlocks change detection, AR/VR and robotics — bilawalsidhu · 2026-11-20
- MRPD replays expert RL logs to consolidate agents at ~1/7 the cost of MOPD — illinois · 2026-11-20
- World Snapshot Models unify 1,000+ visual tasks as generation — World-Snapshot · 2026-11-20
- Multi-agent RL recipe turns pass@k into pass@1 on hard chess puzzles — shizhediao · 2026-11-20
- Steering a Local LLM's Attention Directly: Sliders That Shift Activations, Not Prompts — round · 2026-11-20
- Prime Intellect's RL Residency shipped open evals across robotics, medicine and GPU kernels — lateinteraction · 2026-11-20