Upload SAE blocks.0.attn.hook_z

blocks.0.attn.hook_z/cfg.json

@@ -1 +1 @@
-{"model_name": "tiny-stories-1L-21M", "model_class_name": "HookedTransformer", "hook_name": "blocks.0.attn.hook_z", "hook_eval": "NOT_IN_USE", "hook_layer": 0, "hook_head_index": null, "dataset_path": "georgeyw/dsir-pile-5m-2", "dataset_trust_remote_code": true, "streaming": true, "is_dataset_tokenized": false, "context_size": 512, "use_cached_activations": false, "cached_activations_path": null, "architecture": "standard", "d_in": 256, "d_sae": 4096, "b_dec_init_method": "zeros", "expansion_factor": 16, "activation_fn": "relu", "activation_fn_kwargs": {}, "normalize_sae_decoder": false, "noise_scale": 0.0, "from_pretrained_path": null, "apply_b_dec_to_input": false, "decoder_orthogonal_init": false, "decoder_heuristic_init": true, "init_encoder_as_decoder_transpose": true, "n_batches_in_buffer": 64, "training_tokens": 245760000, "finetuning_tokens": 0, "store_batch_size_prompts": 16, "train_batch_size_tokens": 4096, "normalize_activations": "expected_average_only_in", "device": "cuda", "act_store_device": "cuda", "seed": 42, "dtype": "float32", "prepend_bos": true, "autocast": false, "autocast_lm": false, "compile_llm": false, "llm_compilation_mode": null, "compile_sae": false, "sae_compilation_mode": null, "adam_beta1": 0.9, "adam_beta2": 0.999, "mse_loss_normalization": null, "l1_coefficient": 5, "lp_norm": 1.0, "scale_sparsity_penalty_by_decoder_norm": true, "l1_warm_up_steps": 3000, "lr": 5e-05, "lr_scheduler_name": "constant", "lr_warm_up_steps": 0, "lr_end": 5e-06, "lr_decay_steps": 12000, "n_restart_cycles": 1, "finetuning_method": null, "use_ghost_grads": false, "feature_sampling_window": 1000, "dead_feature_window": 1000, "dead_feature_threshold": 0.0001, "n_eval_batches": 10, "eval_batch_size_prompts": null, "log_to_wandb": false, "log_activations_store_to_wandb": false, "log_optimizer_state_to_wandb": false, "wandb_project": "garrett-sae-lens-test", "wandb_id": null, "run_name": "4096-L1-5-LR-5e-05-Tokens-2.458e+08", "wandb_entity": null, "wandb_log_frequency": 30, "eval_every_n_wandb_logs": 20, "resume": false, "n_checkpoints": 10, "checkpoint_path": "/content/drive/MyDrive/projects/saes/6_real/georgeyw/dsir-pile-5m-2/blocks.0.attn.hook_z/full/tetrahedron/df9tnllg", "verbose": true, "model_kwargs": {}, "model_from_pretrained_kwargs": {"center_writing_weights": false}, "sae_lens_version": "3.21.1", "sae_lens_training_version": "3.21.1", "tokens_per_buffer": 134217728}
+{"model_name": "tiny-stories-1L-21M", "model_class_name": "HookedTransformer", "hook_name": "blocks.0.attn.hook_z", "hook_eval": "NOT_IN_USE", "hook_layer": 0, "hook_head_index": null, "dataset_path": "georgeyw/dsir-pile-5m-2", "dataset_trust_remote_code": true, "streaming": true, "is_dataset_tokenized": false, "context_size": 512, "use_cached_activations": false, "cached_activations_path": null, "architecture": "standard", "d_in": 256, "d_sae": 4096, "b_dec_init_method": "zeros", "expansion_factor": 16, "activation_fn": "relu", "activation_fn_kwargs": {}, "normalize_sae_decoder": false, "noise_scale": 0.0, "from_pretrained_path": null, "apply_b_dec_to_input": false, "decoder_orthogonal_init": false, "decoder_heuristic_init": true, "init_encoder_as_decoder_transpose": true, "n_batches_in_buffer": 64, "training_tokens": 245760000, "finetuning_tokens": 0, "store_batch_size_prompts": 16, "train_batch_size_tokens": 4096, "normalize_activations": "expected_average_only_in", "device": "cuda", "act_store_device": "cuda", "seed": 42, "dtype": "float32", "prepend_bos": true, "autocast": false, "autocast_lm": false, "compile_llm": false, "llm_compilation_mode": null, "compile_sae": false, "sae_compilation_mode": null, "adam_beta1": 0.9, "adam_beta2": 0.999, "mse_loss_normalization": null, "l1_coefficient": 5, "lp_norm": 1.0, "scale_sparsity_penalty_by_decoder_norm": true, "l1_warm_up_steps": 3000, "lr": 5e-05, "lr_scheduler_name": "constant", "lr_warm_up_steps": 0, "lr_end": 5e-06, "lr_decay_steps": 12000, "n_restart_cycles": 1, "finetuning_method": null, "use_ghost_grads": false, "feature_sampling_window": 1000, "dead_feature_window": 1000, "dead_feature_threshold": 0.0001, "n_eval_batches": 10, "eval_batch_size_prompts": null, "log_to_wandb": false, "log_activations_store_to_wandb": false, "log_optimizer_state_to_wandb": false, "wandb_project": "garrett-sae-lens-test", "wandb_id": null, "run_name": "4096-L1-5-LR-5e-05-Tokens-2.458e+08", "wandb_entity": null, "wandb_log_frequency": 30, "eval_every_n_wandb_logs": 20, "resume": false, "n_checkpoints": 10, "checkpoint_path": "../../seed_1/georgeyw/dsir-pile-5m-2/blocks.0.attn.hook_z/full/tetrahedron/eagfcrji", "verbose": true, "model_kwargs": {}, "model_from_pretrained_kwargs": {"center_writing_weights": false}, "sae_lens_version": "3.23.0", "sae_lens_training_version": "3.23.0", "tokens_per_buffer": 134217728}

blocks.0.attn.hook_z/sae_weights.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3be04f70a191e1a6f53a1512165ae8092ef6e473c2ed430ad30d227447ce16c2
+oid sha256:735c4e1d957ae1d9bbe9488238beefbbc2dede5759ffacc024f5d11444a246cf
 size 8406320

blocks.0.attn.hook_z/sparsity.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fb1d1aafdc83ce4f64947c3d0c244caa3d2f4b7ded2cb77c6d4766d7ee4dc8ca
+oid sha256:865b46d22010420a0fc31fdf4369a53d8b88467b5fe1f1c22b5a0f6e8179c660
 size 16464