1
2
3
4
5
|
label,perturb_after,sigma,n_samples,candidates,fd_eps,ckpt_root,ckpt_name,clean_acc,mean_sign_exact,mean_sign_token_acc,worst_sign_exact,best_sign_exact,selected_growth_mean,selected_growth_q90,retain_mean_on_clean_success,retain_worst_on_clean_success,rescue_mean_on_clean_fail,rescue_best_on_clean_fail
smoke_baseline,0,0.0,16,4,0.001,/home/yurenh2/rrm/trm/checkpoints/Sudoku-extreme-1k-aug-1000-ACT-torch/pretrain_mlp_t_sudoku_official_gbs768_repro,step_58590,0.9375,0.9375,0.979938268661499,0.9375,0.9375,-3.8856122493743896,-3.8856122493743896,1.0,1.0,0.0,0.0
smoke_baseline,0,0.03,16,4,0.001,/home/yurenh2/rrm/trm/checkpoints/Sudoku-extreme-1k-aug-1000-ACT-torch/pretrain_mlp_t_sudoku_official_gbs768_repro,step_58590,0.9375,0.875,0.9475308656692505,0.8125,0.9375,-3.8856122493743896,-3.8856122493743896,0.9333333373069763,0.8666666746139526,0.0,0.0
smoke_baseline,8,0.0,16,4,0.001,/home/yurenh2/rrm/trm/checkpoints/Sudoku-extreme-1k-aug-1000-ACT-torch/pretrain_mlp_t_sudoku_official_gbs768_repro,step_58590,1.0,1.0,1.0,1.0,1.0,-5.557522296905518,1.0598664283752441,1.0,1.0,nan,nan
smoke_baseline,8,0.03,16,4,0.001,/home/yurenh2/rrm/trm/checkpoints/Sudoku-extreme-1k-aug-1000-ACT-torch/pretrain_mlp_t_sudoku_official_gbs768_repro,step_58590,1.0,0.9375,0.9764660596847534,0.9375,0.9375,-5.557522296905518,1.0598664283752441,0.9375,0.9375,nan,nan
|
