google-deepmind/physics-IQ-benchmark
Benchmarking physical understanding in generative video models
A star delta has been computed for 62,678 of 63,382 tracked repositories in this period, and only those can be ranked in it.
A repository missing here has not been measured far enough back yet. That does not mean it did not move.
Star counts measured between 02:21 and 17:08 (BKK).
Benchmarking physical understanding in generative video models
Benchmarking synthetic data generation methods.
A Python library for simulation-based inference with deep learning
That's all 3 repositories in this period.