Qwen3-8B-Physics-GRPO-TR/results/validation_mean16.csv

1	step	val_mean16
2	10	0.58359375
3	20	0.59921875
4	30	0.6046875
5	40	0.62109375
6	50	0.65390625
7	60	0.671875
8	70	0.68125
9	80	0.7125
10	90	0.72265625
11	100	0.7296875