RLVR-World: Training World Models with Reinforcement Learning
Paper
•
2505.13934
•
Published
•
14
RLVR-World
-
-
thuml/rt1-frame-tokenizer
Updated • 14 -
thuml/rt1-world-model-single-step-base
0.1B • Updated • 4 -
thuml/rt1-world-model-single-step-rlvr
Updated • 6 -
thuml/rt1-compressive-tokenizer
Updated • 7 -
thuml/rt1-world-model-multi-step-base
0.1B • Updated • 4 -
thuml/rt1-world-model-multi-step-rlvr
0.1B • Updated • 3 -
thuml/webarena-world-model-cot
Viewer • Updated • 6.41k • 47 -
thuml/webarena-world-model-sft
2B • Updated • 5 -
thuml/webarena-world-model-rlvr
2B • Updated • 3 -
thuml/bytesized32-world-model-cot
Viewer • Updated • 304k • 24 -
thuml/bytesized32-world-model-sft
2B • Updated • 4 -
thuml/bytesized32-world-model-rlvr-binary-reward
2B • Updated • 3 -
thuml/bytesized32-world-model-rlvr-task-specific-reward
2B • Updated • 2