sae-lens 6.16.0__py3-none-any.whl → 6.16.3__py3-none-any.whl

sae_lens/__init__.py

@@ -1,5 +1,5 @@
 # ruff: noqa: E402
-__version__ = "6.16.0"
+__version__ = "6.16.3"
 
 import logging
 

sae_lens/saes/batchtopk_sae.py

@@ -23,7 +23,9 @@ class BatchTopK(nn.Module):
     def forward(self, x: torch.Tensor) -> torch.Tensor:
         acts = x.relu()
         flat_acts = acts.flatten()
-        acts_topk_flat = torch.topk(flat_acts, int(self.k * acts.shape[0]), dim=-1)
+        # Calculate total number of samples across all non-feature dimensions
+        num_samples = acts.shape[:-1].numel()
+        acts_topk_flat = torch.topk(flat_acts, int(self.k * num_samples), dim=-1)
         return (
             torch.zeros_like(flat_acts)
             .scatter(-1, acts_topk_flat.indices, acts_topk_flat.values)

sae_lens/saes/matryoshka_batchtopk_sae.py

@@ -78,14 +78,11 @@ class MatryoshkaBatchTopKTrainingSAE(BatchTopKTrainingSAE):
     @override
     def training_forward_pass(self, step_input: TrainStepInput) -> TrainStepOutput:
         base_output = super().training_forward_pass(step_input)
-        hidden_pre = base_output.hidden_pre
         inv_W_dec_norm = 1 / self.W_dec.norm(dim=-1)
         # the outer matryoshka level is the base SAE, so we don't need to add an extra loss for it
         for width in self.cfg.matryoshka_widths[:-1]:
-            inner_hidden_pre = hidden_pre[:, :width]
-            inner_feat_acts = self.activation_fn(inner_hidden_pre)
             inner_reconstruction = self._decode_matryoshka_level(
-                inner_feat_acts, width, inv_W_dec_norm
+                base_output.feature_acts, width, inv_W_dec_norm
             )
             inner_mse_loss = (
                 self.mse_loss_fn(inner_reconstruction, step_input.sae_in)
@@ -105,16 +102,17 @@ class MatryoshkaBatchTopKTrainingSAE(BatchTopKTrainingSAE):
         """
         Decodes feature activations back into input space for a matryoshka level
         """
+        inner_feature_acts = feature_acts[:, :width]
         # Handle sparse tensors using efficient sparse matrix multiplication
         if self.cfg.rescale_acts_by_decoder_norm:
             # need to multiply by the inverse of the norm because division is illegal with sparse tensors
-            feature_acts = feature_acts * inv_W_dec_norm[:width]
-        if feature_acts.is_sparse:
+            inner_feature_acts = inner_feature_acts * inv_W_dec_norm[:width]
+        if inner_feature_acts.is_sparse:
             sae_out_pre = (
-                _sparse_matmul_nd(feature_acts, self.W_dec[:width]) + self.b_dec
+                _sparse_matmul_nd(inner_feature_acts, self.W_dec[:width]) + self.b_dec
             )
         else:
-            sae_out_pre = feature_acts @ self.W_dec[:width] + self.b_dec
+            sae_out_pre = inner_feature_acts @ self.W_dec[:width] + self.b_dec
         sae_out_pre = self.run_time_activation_norm_fn_out(sae_out_pre)
         return self.reshape_fn_out(sae_out_pre, self.d_head)
 
@@ -137,7 +135,3 @@ def _validate_matryoshka_config(cfg: MatryoshkaBatchTopKTrainingSAEConfig) -> No
         warnings.warn(
             "WARNING: You have only set one matryoshka level. This is equivalent to using a standard BatchTopK SAE and is likely not what you want."
         )
-    if cfg.matryoshka_widths[0] < cfg.k:
-        raise ValueError(
-            "The smallest matryoshka level width cannot be smaller than cfg.k."
-        )

{sae_lens-6.16.0.dist-info → sae_lens-6.16.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sae-lens
-Version: 6.16.0
+Version: 6.16.3
 Summary: Training and Analyzing Sparse Autoencoders (SAEs)
 License: MIT
 License-File: LICENSE

{sae_lens-6.16.0.dist-info → sae_lens-6.16.3.dist-info}/RECORD

@@ -1,4 +1,4 @@
-sae_lens/__init__.py,sha256=8-kJQbVwfBSm9fla6sxHlPwLxQ9ghneJanQJVL1i_gc,3886
+sae_lens/__init__.py,sha256=c1rxG64QCdP4n1LI8Du_dxEn30E1fXXbEfZ0kaZ2JiI,3886
 sae_lens/analysis/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 sae_lens/analysis/hooked_sae_transformer.py,sha256=vRu6JseH1lZaEeILD5bEkQEQ1wYHHDcxD-f2olKmE9Y,14275
 sae_lens/analysis/neuronpedia_integration.py,sha256=Gx1W7hUBEuMoasNcnOnZ1wmqbXDd1pSZ1nqKEya1HQc,4962
@@ -15,10 +15,10 @@ sae_lens/pretokenize_runner.py,sha256=x-reJzVPFDS9iRFbZtrFYSzNguJYki9gd0pbHjYJ3r
 sae_lens/pretrained_saes.yaml,sha256=6ca3geEB6NyhULUrmdtPDK8ea0YdpLp8_au78vIFC5w,602553
 sae_lens/registry.py,sha256=nhy7BPSudSATqW4lo9H_k3Na7sfGHmAf9v-3wpnLL_o,1490
 sae_lens/saes/__init__.py,sha256=sIfZUxZ4m3HPtPymCJlpBEofiCrL8_QziE6ChS-v4lE,1677
-sae_lens/saes/batchtopk_sae.py,sha256=zxIke8lOBKkQEMVFk6sSW6q_s6F9RKhysLqfqG9ecwI,5300
+sae_lens/saes/batchtopk_sae.py,sha256=x4EbgZl0GUickRPcCmtKNGS2Ra3Uy1Z1OtF2FnrSabQ,5422
 sae_lens/saes/gated_sae.py,sha256=qcmM9JwBA8aZR8z_IRHV1_gQX-q_63tKewWXRnhdXuo,8986
 sae_lens/saes/jumprelu_sae.py,sha256=HHBF1sJ95lZvxwP5vwLSQFKdnJN2KKYK0WAEaLTrta0,13399
-sae_lens/saes/matryoshka_batchtopk_sae.py,sha256=zrS4MksbxdhhftmU3UWjRCWjR7iEBpAk6N00c6GrXks,6291
+sae_lens/saes/matryoshka_batchtopk_sae.py,sha256=4_1cVaxk6c6jgJEbxqebtG-cjQNIzaMAfjSPGfR7_VU,6062
 sae_lens/saes/sae.py,sha256=vABlwyZ0JtL896xxBGIoqfiByoszIf-e4ggPgz34RL0,38300
 sae_lens/saes/standard_sae.py,sha256=9UqYyYtQuThYxXKNaDjYcyowpOx2-7cShG-TeUP6JCQ,5940
 sae_lens/saes/topk_sae.py,sha256=tzQM5eQFifMe--8_8NUBYWY7hpjQa6A_olNe6U71FE8,21275
@@ -34,7 +34,7 @@ sae_lens/training/types.py,sha256=1FpLx_Doda9vZpmfm-x1e8wGBYpyhe9Kpb_JuM5nIFM,90
 sae_lens/training/upload_saes_to_huggingface.py,sha256=r_WzI1zLtGZ5TzAxuG3xa_8T09j3zXJrWd_vzPsPGkQ,4469
 sae_lens/tutorial/tsea.py,sha256=fd1am_XXsf2KMbByDapJo-2qlxduKaa62Z2qcQZ3QKU,18145
 sae_lens/util.py,sha256=tCovQ-eZa1L7thPpNDL6PGOJrIMML2yLI5e0EHCOpS8,3309
-sae_lens-6.16.0.dist-info/METADATA,sha256=VI8vndYrn1pKxiHUDjyBNx74_S8mHwiaTK1nfo7t0BU,5318
-sae_lens-6.16.0.dist-info/WHEEL,sha256=zp0Cn7JsFoX2ATtOhtaFYIiE2rmFAD4OcMhtUki8W3U,88
-sae_lens-6.16.0.dist-info/licenses/LICENSE,sha256=DW6e-hDosiu4CfW0-imI57sV1I5f9UEslpviNQcOAKs,1069
-sae_lens-6.16.0.dist-info/RECORD,,
+sae_lens-6.16.3.dist-info/METADATA,sha256=sZBPyQRO8rpyB21LtAfOtfxiXhBQ7RiG8F3IluD_8mw,5318
+sae_lens-6.16.3.dist-info/WHEEL,sha256=zp0Cn7JsFoX2ATtOhtaFYIiE2rmFAD4OcMhtUki8W3U,88
+sae_lens-6.16.3.dist-info/licenses/LICENSE,sha256=DW6e-hDosiu4CfW0-imI57sV1I5f9UEslpviNQcOAKs,1069
+sae_lens-6.16.3.dist-info/RECORD,,