task stringclasses 1
value | env_id stringclasses 1
value | env_kwargs dict | upstream_env_id stringclasses 1
value | seed int64 2.03B 2.33B | episode_seed int64 2.03B 2.33B | extra_state null | timestamp stringdate 2026-04-28 20:47:12 2026-04-28 21:10:28 | history listlengths 2 20 | stats dict | init_state dict | init_state_hash stringlengths 64 64 | hash stringlengths 64 64 | solve_kwargs dict | dataset_pipeline stringclasses 1
value | version stringclasses 1
value | episode int64 0 6.74k | difficulty stringclasses 1
value | generated_at stringdate 2026-04-28 20:47:12 2026-04-28 21:10:28 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
sokoban_medium | sokoban/hard | {"dim_room":[10,10],"max_steps":40,"num_boxes":3,"num_gen_steps":34,"success_mode":"stop_then_solved(...TRUNCATED) | Sokoban-v1-3box | 2,026,042,800 | 2,026,042,800 | null | 2026-04-28T20:47:12.728320 | [{"step":0,"prompt":"You are playing Sokoban, a box-pushing puzzle. Your job is to control the green(...TRUNCATED) | {"step":12,"reward":0.0,"terminated":true,"truncated":false,"rollout_capped":false,"rollout_max_step(...TRUNCATED) | {"room_fixed":[[0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0],[0,0,1,1,1,2,1,1,1,0],[0,1,1,1,2,1,1,1,1,(...TRUNCATED) | bf5e963a5232f40bf9bac2eb056eb0459424eca69cfdba52505adfeca0a5c780 | bf5e963a5232f40bf9bac2eb056eb0459424eca69cfdba52505adfeca0a5c780 | {
"policy": "random",
"random_stop_prob": 0.05,
"stop_eligible_after_step": 0
} | random_policy_world_model | sokoban_random_policy_world_model_v1 | 0 | medium | 2026-04-28T20:47:12.728491 |
sokoban_medium | sokoban/hard | {"dim_room":[10,10],"max_steps":40,"num_boxes":3,"num_gen_steps":34,"success_mode":"stop_then_solved(...TRUNCATED) | Sokoban-v1-3box | 2,026,042,800 | 2,026,042,801 | null | 2026-04-28T20:47:12.869182 | [{"step":0,"prompt":"You are playing Sokoban, a box-pushing puzzle. Your job is to control the green(...TRUNCATED) | {"step":2,"reward":0.0,"terminated":true,"truncated":false,"rollout_capped":false,"rollout_max_steps(...TRUNCATED) | {"room_fixed":[[0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0],[0,1,1,1,0,0,0,0,0,(...TRUNCATED) | 355a05cd17f1a9c9f62fbcd123b3edd21e3ffbfb1138325f6b0802315f767825 | 355a05cd17f1a9c9f62fbcd123b3edd21e3ffbfb1138325f6b0802315f767825 | {
"policy": "random",
"random_stop_prob": 0.05,
"stop_eligible_after_step": 0
} | random_policy_world_model | sokoban_random_policy_world_model_v1 | 1 | medium | 2026-04-28T20:47:12.869417 |
sokoban_medium | sokoban/hard | {"dim_room":[10,10],"max_steps":40,"num_boxes":3,"num_gen_steps":34,"success_mode":"stop_then_solved(...TRUNCATED) | Sokoban-v1-3box | 2,026,042,800 | 2,026,042,802 | null | 2026-04-28T20:47:13.736728 | [{"step":0,"prompt":"You are playing Sokoban, a box-pushing puzzle. Your job is to control the green(...TRUNCATED) | {"step":20,"reward":0.0,"terminated":false,"truncated":false,"rollout_capped":true,"rollout_max_step(...TRUNCATED) | {"room_fixed":[[0,0,0,0,0,0,0,0,0,0],[0,1,1,0,0,0,0,0,0,0],[0,1,1,1,0,0,0,0,0,0],[0,2,1,1,2,0,0,0,0,(...TRUNCATED) | c7ac453bbd36b5071087dba5b6c555f3aa1cf7e244fe407bfa1056b6ea07f55a | c7ac453bbd36b5071087dba5b6c555f3aa1cf7e244fe407bfa1056b6ea07f55a | {
"policy": "random",
"random_stop_prob": 0.05,
"stop_eligible_after_step": 0
} | random_policy_world_model | sokoban_random_policy_world_model_v1 | 2 | medium | 2026-04-28T20:47:13.736922 |
sokoban_medium | sokoban/hard | {"dim_room":[10,10],"max_steps":40,"num_boxes":3,"num_gen_steps":34,"success_mode":"stop_then_solved(...TRUNCATED) | Sokoban-v1-3box | 2,026,042,800 | 2,026,042,803 | null | 2026-04-28T20:47:14.207635 | [{"step":0,"prompt":"You are playing Sokoban, a box-pushing puzzle. Your job is to control the green(...TRUNCATED) | {"step":10,"reward":0.0,"terminated":true,"truncated":false,"rollout_capped":false,"rollout_max_step(...TRUNCATED) | {"room_fixed":[[0,0,0,0,0,0,0,0,0,0],[0,2,1,1,1,1,1,1,1,0],[0,1,0,0,0,0,1,1,1,0],[0,1,1,0,0,1,1,2,1,(...TRUNCATED) | ce775bb84ed2112900a36a74d72e6e749d83bd942231db115329a218965be4e0 | ce775bb84ed2112900a36a74d72e6e749d83bd942231db115329a218965be4e0 | {
"policy": "random",
"random_stop_prob": 0.05,
"stop_eligible_after_step": 0
} | random_policy_world_model | sokoban_random_policy_world_model_v1 | 3 | medium | 2026-04-28T20:47:14.208008 |
sokoban_medium | sokoban/hard | {"dim_room":[10,10],"max_steps":40,"num_boxes":3,"num_gen_steps":34,"success_mode":"stop_then_solved(...TRUNCATED) | Sokoban-v1-3box | 2,026,042,800 | 2,026,042,804 | null | 2026-04-28T20:47:14.710312 | [{"step":0,"prompt":"You are playing Sokoban, a box-pushing puzzle. Your job is to control the green(...TRUNCATED) | {"step":10,"reward":0.0,"terminated":true,"truncated":false,"rollout_capped":false,"rollout_max_step(...TRUNCATED) | {"room_fixed":[[0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,1,1,0],[0,1,0,0,0,0,0,1,1,0],[0,2,1,1,0,0,0,0,1,(...TRUNCATED) | 3eae8547c06dc1c9b35bdc3e56c5f1af28fd56bb1ac31be891fc461774be084a | 3eae8547c06dc1c9b35bdc3e56c5f1af28fd56bb1ac31be891fc461774be084a | {
"policy": "random",
"random_stop_prob": 0.05,
"stop_eligible_after_step": 0
} | random_policy_world_model | sokoban_random_policy_world_model_v1 | 4 | medium | 2026-04-28T20:47:14.710546 |
sokoban_medium | sokoban/hard | {"dim_room":[10,10],"max_steps":40,"num_boxes":3,"num_gen_steps":34,"success_mode":"stop_then_solved(...TRUNCATED) | Sokoban-v1-3box | 2,026,042,800 | 2,026,042,805 | null | 2026-04-28T20:47:15.230648 | [{"step":0,"prompt":"You are playing Sokoban, a box-pushing puzzle. Your job is to control the green(...TRUNCATED) | {"step":11,"reward":0.0,"terminated":true,"truncated":false,"rollout_capped":false,"rollout_max_step(...TRUNCATED) | {"room_fixed":[[0,0,0,0,0,0,0,0,0,0],[0,1,1,0,0,0,0,0,0,0],[0,1,1,2,0,0,0,0,0,0],[0,1,1,1,0,0,0,0,0,(...TRUNCATED) | cc66c15edee4d55985b7ac67afe4e1bf3f561f330747e16a592af26d25c6e18c | cc66c15edee4d55985b7ac67afe4e1bf3f561f330747e16a592af26d25c6e18c | {
"policy": "random",
"random_stop_prob": 0.05,
"stop_eligible_after_step": 0
} | random_policy_world_model | sokoban_random_policy_world_model_v1 | 5 | medium | 2026-04-28T20:47:15.230907 |
sokoban_medium | sokoban/hard | {"dim_room":[10,10],"max_steps":40,"num_boxes":3,"num_gen_steps":34,"success_mode":"stop_then_solved(...TRUNCATED) | Sokoban-v1-3box | 2,026,042,800 | 2,026,042,806 | null | 2026-04-28T20:47:15.752451 | [{"step":0,"prompt":"You are playing Sokoban, a box-pushing puzzle. Your job is to control the green(...TRUNCATED) | {"step":11,"reward":0.0,"terminated":true,"truncated":false,"rollout_capped":false,"rollout_max_step(...TRUNCATED) | {"room_fixed":[[0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,2,1,1,1,0],[0,0,0,0,0,1,1,1,1,0],[0,0,0,0,0,1,1,1,1,(...TRUNCATED) | 3dcd4654c7f3a42a0ee6141758b7ddd7014be61180eea8f1a25fa86239d03b84 | 3dcd4654c7f3a42a0ee6141758b7ddd7014be61180eea8f1a25fa86239d03b84 | {
"policy": "random",
"random_stop_prob": 0.05,
"stop_eligible_after_step": 0
} | random_policy_world_model | sokoban_random_policy_world_model_v1 | 6 | medium | 2026-04-28T20:47:15.752704 |
sokoban_medium | sokoban/hard | {"dim_room":[10,10],"max_steps":40,"num_boxes":3,"num_gen_steps":34,"success_mode":"stop_then_solved(...TRUNCATED) | Sokoban-v1-3box | 2,026,042,800 | 2,026,042,807 | null | 2026-04-28T20:47:16.261193 | [{"step":0,"prompt":"You are playing Sokoban, a box-pushing puzzle. Your job is to control the green(...TRUNCATED) | {"step":10,"reward":0.0,"terminated":true,"truncated":false,"rollout_capped":false,"rollout_max_step(...TRUNCATED) | {"room_fixed":[[0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0],[0,1,0,0,0,1,0,0,0,0],[0,1,1,1,1,1,0,0,0,(...TRUNCATED) | cc2ef772bbeb8e0d0fe8d3e3b51af96d8fda0818660896f498dea7f37c5e8f8e | cc2ef772bbeb8e0d0fe8d3e3b51af96d8fda0818660896f498dea7f37c5e8f8e | {
"policy": "random",
"random_stop_prob": 0.05,
"stop_eligible_after_step": 0
} | random_policy_world_model | sokoban_random_policy_world_model_v1 | 7 | medium | 2026-04-28T20:47:16.261783 |
sokoban_medium | sokoban/hard | {"dim_room":[10,10],"max_steps":40,"num_boxes":3,"num_gen_steps":34,"success_mode":"stop_then_solved(...TRUNCATED) | Sokoban-v1-3box | 2,026,042,800 | 2,026,042,808 | null | 2026-04-28T20:47:17.111518 | [{"step":0,"prompt":"You are playing Sokoban, a box-pushing puzzle. Your job is to control the green(...TRUNCATED) | {"step":20,"reward":0.0,"terminated":false,"truncated":false,"rollout_capped":true,"rollout_max_step(...TRUNCATED) | {"room_fixed":[[0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,(...TRUNCATED) | d7907dfde7e3ddffc2de22a114edc51cbd51a1128977fbd084ac7c0724190ef2 | d7907dfde7e3ddffc2de22a114edc51cbd51a1128977fbd084ac7c0724190ef2 | {
"policy": "random",
"random_stop_prob": 0.05,
"stop_eligible_after_step": 0
} | random_policy_world_model | sokoban_random_policy_world_model_v1 | 8 | medium | 2026-04-28T20:47:17.111700 |
sokoban_medium | sokoban/hard | {"dim_room":[10,10],"max_steps":40,"num_boxes":3,"num_gen_steps":34,"success_mode":"stop_then_solved(...TRUNCATED) | Sokoban-v1-3box | 2,026,042,800 | 2,026,042,809 | null | 2026-04-28T20:47:17.692261 | [{"step":0,"prompt":"You are playing Sokoban, a box-pushing puzzle. Your job is to control the green(...TRUNCATED) | {"step":14,"reward":0.0,"terminated":true,"truncated":false,"rollout_capped":false,"rollout_max_step(...TRUNCATED) | {"room_fixed":[[0,0,0,0,0,0,0,0,0,0],[0,0,0,2,1,1,1,1,2,0],[0,0,0,1,1,1,2,1,1,0],[0,0,0,0,0,1,1,1,1,(...TRUNCATED) | eff3fb48060d150415773ffaa2d70925203e3ecb1bf5a5d405ab823082527c8a | eff3fb48060d150415773ffaa2d70925203e3ecb1bf5a5d405ab823082527c8a | {
"policy": "random",
"random_stop_prob": 0.05,
"stop_eligible_after_step": 0
} | random_policy_world_model | sokoban_random_policy_world_model_v1 | 9 | medium | 2026-04-28T20:47:17.692541 |
End of preview. Expand in Data Studio
VisGym Sokoban Random Policy World-Model Trajectories
This dataset contains random-policy interaction trajectories for the custom VisGym Sokoban environment.
Default generation layout:
data/sokoban_medium/train/*.jsonl.gz- 20-step rollout cap
- medium task config:
sokoban/hardwith 10x10 room, 3 boxes, reverse-generated levels, and stop-then-solved success semantics
Each row is one environment trajectory. History entries include base64 JPEG frames, prompt,
sampled action, reward, and environment info. For training compatibility, every step has
image_prev == image; the final step also has image_next == image.
The random policy samples valid move/push actions uniformly and samples the valid stop action with the configured random stop probability after the first step.
Generation counts, hash audit, and validation metadata are in metadata/.
Target Hub repo: https://huggingface.co/datasets/novastar112/sokoban_v0_random
- Downloads last month
- 660