Skip to main content
zachdata
Projects
matryoshka-vsae-batch-topk-experiments
Workspace
Log in
Sign up
Overview
Workspace
Runs
Automat.
Sweeps
Reports
Artifacts
Zachdata's workspace
Personal workspace
Automated workspace
Changes are only visible to you.
Runs
13
Name
5 visualized
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr7e-05_kl1_learned_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr7e-05_kl1_learned_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr7e-05_kl1_learned_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr7e-05_kl1_learned_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.007_kl1_learned_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.007_kl1_learned_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.001_kl1_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.001_kl1_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl0.01_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl0.01_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl1.0_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl1.0_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl30.0_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl30.0_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl100.0_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl100.0_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0005_kl500.0_fixed_var_kl_warmup100_k32_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0005_kl500.0_fixed_var_kl_warmup100_k32_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0005_kl500.0_fixed_var_kl_warmup100_k32_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0005_kl500.0_fixed_var_kl_warmup100_k32_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0005_kl500.0_fixed_var_kl_warmup100_k32_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0005_kl500.0_fixed_var_kl_warmup100_k32_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0005_kl500.0_fixed_var_kl_warmup100_k32_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0005_kl500.0_fixed_var_kl_warmup100_k32_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl40.0_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl40.0_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
1-13
of 13
frac_variance_explained
frac_variance_explained
0
10
20
30
40
50
Step
0.2
0.4
0.6
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.001_kl1_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl0.01_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl1.0_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl30.0_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
MatryoshkaVSAE_BatchTopK_gelu-1l_d8192_lr0.0007_kl100.0_fixed_var_kl_warmup1500_k65_magnitude_trainer_0
Previous
Next