Navigation and planning
AtCoder · CodeChef · Gymnasium · Jumanji · MiniGrid
- Collections
- 9
- Task profiles
- 109
Public Benchmark surface
Independently installable Benchmark distributions across control, planning, robotics, driving, and long-horizon games.
Generated from the current catalog.
AtCoder · CodeChef · Gymnasium · Jumanji · MiniGrid
DeepMind Control Suite · Gymnasium · Gymnasium-Robotics · MetaWorld · robosuite
ALE · ARC Prize · Crafter · HighwayEnv · Jackdaw · NLE · Stable-Retro · ViZDoom
Long-horizon constraint placement against a deterministic explorer with private targets.
Schedule low-cost bonds among moving components on a toroidal plane.
Five hundred turns of hierarchical machine investment and compounding production.
Full-range warehouse storage and retrieval through a compact forklift instruction language.
Small state-space control tasks with discrete or scalar continuous Actions.
Rule-based discrete tasks for planning, stochastic transitions, and delayed reward.
Landing, locomotion, and visual driving in medium-weight physics environments.
Continuous control from small arms and pendulums to contact-rich locomotion.
Canonical state-based continuous-control tasks from balancing and locomotion to manipulation.
Partially observable navigation, memory, object interaction, and procedural layouts.
Wave Function Collapse layouts spanning mazes, rooms, dungeons, obstacles, and caves.
Language-conditioned instruction following from atomic navigation to composite missions.
Structured combinatorial environments with exact action masks and task-specific progress diagnostics.
Goal-conditioned reaching, manipulation, maze, dexterous hand, and kitchen tasks.
Single-task and multi-task robotic manipulation collections.
Single- and dual-arm Panda manipulation with dense task state and per-Episode visual evidence.
Traffic negotiation, parking, racing, and lane keeping.
Open-world survival, resource collection, crafting, and long-horizon development from RGB observations.
Interactive visual reasoning across complete games, in-Episode level resets, and official scorecard aggregation.
Long-horizon procedural dungeon survival with complete semantic training trajectories.
Arcade control using the redistributable Tetris ROM.
First-person navigation and combat across the standard packaged scenarios.
Genesis arcade control from the packaged Level 1 state.
Long-horizon deckbuilding with hands, shops, Jokers, consumables, packs, and antes.
Authoring boundary
External distributions use only the public authoring SPI and own their domain semantics, dependencies, Feedback, and conformance tests.
Read the authoring guide →