| {"gsae_repo": "jacobcd52/mats=gpt2-saes", "gsae_filename_no_suffix": "gpt2-small_l1_coeff=45_expansion=64_tokens=409600000_lr=1e-05", "control_dataset_path": null, "is_control_dataset_tokenized": true, "control_mixture": 0.0, "model_name": "gpt2-small", "model_class_name": "HookedTransformer", "hook_name": "blocks.8.hook_resid_pre", "hook_eval": "NOT_IN_USE", "hook_layer": 8, "hook_head_index": null, "dataset_path": "jacobcd52/physics-papers", "dataset_trust_remote_code": true, "streaming": true, "is_dataset_tokenized": false, "context_size": 256, "use_cached_activations": false, "cached_activations_path": null, "architecture": "gated", "d_in": 768, "d_sae": 384, "b_dec_init_method": "zeros", "expansion_factor": 0.5, "activation_fn": "relu", "normalize_sae_decoder": false, "noise_scale": 0.0, "from_pretrained_path": null, "apply_b_dec_to_input": true, "decoder_orthogonal_init": false, "decoder_heuristic_init": true, "init_encoder_as_decoder_transpose": true, "n_batches_in_buffer": 64, "training_tokens": 24576000, "finetuning_tokens": 0, "store_batch_size_prompts": 16, "train_batch_size_tokens": 8192, "normalize_activations": "none", "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": 6, "lp_norm": 1.0, "scale_sparsity_penalty_by_decoder_norm": false, "l1_warm_up_steps": 150, "lr": 0.001, "lr_scheduler_name": "constant", "lr_warm_up_steps": 0, "lr_end": 0.0001, "lr_decay_steps": 600, "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": true, "log_activations_store_to_wandb": false, "log_optimizer_state_to_wandb": false, "wandb_project": "phys-SSAE-wide-gsae", "wandb_id": null, "run_name": "l1=6_expansion=0.5_control_mix=0.0_tokens=24576000_lr=0.001", "wandb_entity": null, "wandb_log_frequency": 30, "eval_every_n_wandb_logs": 20, "resume": false, "n_checkpoints": 0, "checkpoint_path": "gpt2_gsae_l1_coeff=6_expansion=0.5_tokens=24576000_lr=0.001/30ds4e5j", "verbose": true, "model_kwargs": {}, "model_from_pretrained_kwargs": {}, "sae_lens_version": "3.7.0", "sae_lens_training_version": "3.7.0", "tokens_per_buffer": 134217728} |