interpretune.analysis.ops.definitions
=====================================

.. automodule:: interpretune.analysis.ops.definitions

   
   .. rubric:: Functions

   .. autosummary::
   
      ablate_sae_latent
      ablation_attribution_impl
      boolean_logits_to_avg_logit_diff
      compute_attribution_graph_impl
      concept_direction_impl
      extract_concept_latent_examples_impl
      extract_concept_latent_state_impl
      extract_top_features_impl
      feature_intervention_forward_impl
      get_alive_latents_impl
      get_answer_indices_impl
      get_loss_preds_diffs
      gradient_attribution_impl
      graph_node_influence_impl
      graph_prune_impl
      labels_to_ids_impl
      logit_diffs_impl
      model_ablation_impl
      model_cache_forward_impl
      model_forward_impl
      model_fwd_impl
      model_fwd_intervention_impl
      model_fwd_w_cache_impl
      model_fwd_w_cache_latent_models_impl
      model_gradient_impl
      sae_correct_acts_impl