torchlens 2.26.0

A package for extracting activations from PyTorch models and saving portable bundles

TorchLens

See, save, and steer any PyTorch model. TorchLens captures every activation and gradient -- across the forward and backward pass -- auto-visualizes the full computational graph, exposes rich per-op metadata, and lets you intervene on the network as it runs. Any architecture, even dynamic and recurrent ones.

Tested on over 700 models (image, video, audio, multimodal, language; feedforward, recurrent, transformer, GNN) — and it records every last detail of every part of your model: 180+ metadata fields per operation, and 550+ fields in total across every record type — operations, modules, parameters, buffers, gradients, and the model itself.

import torch, torchvision.models as models, torchlens as tl

model = models.alexnet(weights=None)
x = torch.randn(1, 3, 224, 224)

log = tl.trace(model, x)     # one call -- full graph + all activations
print(log.summary())          # module table, op count, FLOPs
print(log['relu_1_2'].out.shape)   # grab any activation by name ...
print(log['features.6'].out.shape) # ... or by module path
print(log[7].func_name)            # ... or by ordinal
log.draw()                    # PDF of the computational graph

Quick Links

• Paper | 10-minute tutorial notebook | Facets tutorial | 5-minute gallery | 50-minute gallery
• Performance guide | AI-agent quick reference | Limitations | Migration tables

Installation

Install Graphviz first (required for graph visualizations), then TorchLens:

sudo apt install graphviz   # Debian/Ubuntu; see graphviz.org for other platforms
pip install torchlens

Compatible with PyTorch 1.8.0+.

Quickstart

import torch
import torchvision.models as models
import torchlens as tl

model = models.alexnet(weights=None)
x = torch.randn(1, 3, 224, 224)

log = tl.trace(model, x)
print(log.summary())

Model: AlexNet
+-----------------------------+---------------+--------+-------+
| Layer                       | Output Shape  | Params | Train |
+-----------------------------+---------------+--------+-------+
| input                       | [1,3,224,224] | 0      | -     |
| features (Sequential)       | [1,256,6,6]   | 2.5 M  | yes   |
| avgpool (AdaptiveAvgPool2d) | [1,256,6,6]   | 0      | -     |
| classifier (Sequential)     | [1,1000]      | 58.6 M | yes   |
| output                      | [1,1000]      | -      | -     |
+-----------------------------+---------------+--------+-------+
Params: 61,100,840 unique; trainable: 61,100,840
Ops: 22 total
Edges: 23 total
Forward FLOPs: 1.4 GFLOPs  MACs: 718.9 MFLOPs

Index any operation by name, module path, or ordinal:

log['relu_1_2'].out.shape      # torch.Size([1, 64, 55, 55])
log['features.6'].out.shape    # same op via module path
log[7].func_name               # 'conv2d'
log['conv2d_3'].out.shape      # short name (ordinal suffix optional)
log[-1].layer_label            # 'output_1'

Visualize the graph as a PDF:

log.draw()                        # unrolled by default
log.draw(vis_mode='rolled')       # rolled (compact for recurrent)
log.draw(vis_mode='unrolled')     # every pass as a distinct node

What You Can Do

1. Flexible feature extraction

Save everything, or select exactly what you need:

# Save only relu activations
log = tl.trace(model, x, save=tl.func('relu'))

# Save all ops inside the 'encoder' submodule
log = tl.trace(model, x, save=tl.in_module('encoder'))

# Save conv2d ops that are immediately followed by a relu, keeping a 4-op lookback window
conv_before_relu = tl.func('conv2d') & tl.followed_by(tl.func('relu'))
log = tl.trace(model, x, save=conv_before_relu,
               lookback=4, lookback_payload_policy='detached_raw')

# Stop capture early (can be faster than a plain forward pass)
log = tl.trace(model, x, save=tl.in_module('layer2'), halt=tl.in_module('layer2'))

# Lightweight sparse recording for tight loops -- materialize structure later
recording = tl.record(model, x, save=tl.func('relu'))
trace = recording.to_trace()

# One-line activation pull
act = tl.pluck(model, x, 'relu_1_2')   # returns tensor directly

# Batch extraction across a dataset
tl.extract_dataset(model, dataset, layers=['relu_1_2', 'conv2d_3_7'],
                   batch_size=32, output_dir='activations/')

Performance note: With halt= and tl.record, capture can run faster than the raw forward pass -- measured at 0.84x raw on ResNet-18 and 0.83x on GPT-2 (HookedTransformer) at 25% depth. Full exhaustive capture runs at roughly 14x the raw forward and amortizes on large models. See docs/performance.md for the full benchmark table.

Save and load traces portably:

tl.save(log, 'my_trace')
loaded = tl.load('my_trace')

2. Forward AND backward pass

Capture per-op gradients with the same API:

x = torch.randn(1, 3, 224, 224, requires_grad=True)
log = tl.trace(model, x, save_grads=True)
log.log_backward(log[log.output_layers[0]].out.sum())

grad = log['relu_1_2'].grad      # gradient tensor flowing through that op
print(grad.shape)                 # torch.Size([1, 64, 55, 55])

Narrow gradient saving to specific ops with the same selector predicates:

log = tl.trace(model, x, save_grads=tl.func('relu'))
log.log_backward(log[log.output_layers[0]].out.sum())

Backward capture is PyTorch-only. Non-torch backends expose derived leaf-level gradients through a second AD pass. See docs/backward.md.

3. Vast metadata per operation

Every operation records shape, dtype, device, timing, FLOPs, parameter info, module containment, graph distances, conditional context, RNG state, and more. The full print of any op includes all of this:

print(log['conv2d_3_7'])

Layer conv2d_3_7, operation 7/22:
    Output tensor: shape=(1, 384, 13, 13), dtype=torch.float32, size=253.5 KB
        tensor([[-0.0198,  0.0946,  0.1109, ...
    Related Layers:
        - parent layers: maxpool2d_2_6
        - child layers: relu_3_8
    Params: Computed from params with shape (384, 192, 3, 3), (384,); 663936 params total (2.5 MB)
    Function: conv2d (grad_fn_handle: ConvolutionBackward0)
    Computed inside module: features.6:1
    Config: out_channels=384, in_channels=192, kernel_size=(3, 3), padding=(1, 1)
    Time elapsed: 1.4 ms
    Lookup keys: -17, 7, conv2d_3, conv2d_3:1, conv2d_3_7, conv2d_3_7:1, features.6, features.6:1

Every op also records the Python call stack that produced it, with file and line number:

loc = log['conv2d_3_7'].code_context[0]
print(loc.file, loc.line_number, loc.func_name)

Metadata is available as pandas DataFrames:

df = log.to_pandas()            # one row per op
params_df = log.params.to_pandas()
modules_df = log.modules.to_pandas()

4. Automatic visualization

log.draw()                           # default: unrolled with sibling ordering
log.draw(vis_mode='rolled')          # compact rolled layout
log.draw(vis_mode='unrolled')        # every pass as a distinct node

Control nesting depth to zoom in on submodules:

For recurrent models, the rolled view collapses repeated structure cleanly:

class SimpleRecurrent(torch.nn.Module):
    def __init__(self):
        super().__init__()
        self.fc = torch.nn.Linear(in_features=5, out_features=5)
    def forward(self, x):
        for r in range(4):
            x = self.fc(x)
            x = x + 1
            x = x * 2
        return x

model = SimpleRecurrent()
x = torch.randn(6, 5)
log = tl.trace(model, x)
print(log['linear_1:2'].out)     # second pass of the linear layer
log.draw(vis_mode='rolled')

5. Interventions

Ablate, steer, scale, or replace activations during the forward pass:

# Zero-ablate all relu activations inline during capture
ablated = tl.trace(model, x, save=tl.func('relu'),
                   intervene=tl.when(tl.func('relu'), tl.zero_ablate()))
print(ablated['relu_1_2'].out.abs().max())  # tensor(0.)

# Scale relus to 50%
scaled = tl.trace(model, x, save=tl.func('relu'),
                  intervene=tl.when(tl.func('relu'), tl.scale(0.5)))

Available helpers: tl.zero_ablate, tl.mean_ablate, tl.resample_ablate, tl.steer, tl.scale, tl.clamp, tl.noise, tl.project_onto, tl.project_off, tl.swap_with, tl.splice_module.

For post-hoc DAG replay and isolated experiments, capture with intervention_ready=True and use log.fork() + log.replay() / log.rerun(model, x). Live hooks during rerun require capture-time selectors (e.g. tl.func(...), tl.module(...)); finalized labels resolve via log.find_sites(...). See docs/intervention_api.md for the full reference.

Compare multiple runs side by side with tl.bundle:

bundle = tl.bundle({'clean': clean_log, 'patched': patched_log}, baseline='clean')
bundle.compare_at(tl.func('relu'))

Facets provide named sub-views for attention heads, LSTM outputs, and fused projections (for models with those structures):

# ViT / transformer model with attention blocks
log = tl.trace(vit_model, x)
q = log.modules['blocks.0.attn'].facets['q']    # query vectors for head 0
h_n = log.modules['lstm'].facets['h_n']         # LSTM final hidden state

See docs/facets.md for the full facets reference, including activation patching helpers, SDPA reconstruction, and TransformerLens aliases.

See docs/intervention_api.md for the full selector and helper reference.

6. Works on anything, including dynamic and recurrent models

TorchLens uses eager-mode Python-level function wrapping rather than graph tracing. This means it captures whatever actually runs, including:

• Dynamic control flow (if/else branching, loops, early exits)
• Recurrent architectures (RNNs, LSTMs, state-space models)
• Transformer variants including fused attention
• Graph neural networks
• Mixed architectures

This is the key differentiator from static-graph extractors like torchvision.feature_extraction, which require static computational graphs and cannot handle dynamic architectures.

Multi-backend. The same tl.trace API works across frameworks via backend=:

Capability	PyTorch	JAX (preview)	tinygrad (preview)	MLX (preview)	Paddle (preview)	TensorFlow (preview)
Forward capture + graph/metadata	yes	yes	yes	yes	yes	yes
Module hierarchy	torch_module	Equinox/Flax NNX pytree_module; raw function_root	object_module; raw function_root	object_module; raw function_root	object_module; raw function_root	Keras/tf.Module object_module; raw function_root
Control-flow unroll	eager Python	lax.scan/cond/while_loop	lazy UOp graph	limited	dygraph/eager Python only	eager Python control flow
Static-label save=	yes	yes	yes	yes	yes	yes
Portable array .tlspec payloads	full	forward/derived arrays	forward/derived arrays	forward/derived arrays	forward/derived arrays	forward arrays
Gradients	full backward graph	leaf-level + zero-tap T1 intermediate derived	leaf-level + T1 intermediate derived	leaf-level + custom-VJP-tap T1 intermediate derived	leaf-level + T1 intermediate derived	deferred
Interventions / halt / fastlog	yes	--	--	--	--	--

log = tl.trace(torch_model, x)                      # PyTorch (default)
log = tl.trace(jax_fn,      inputs, backend='jax')  # JAX preview
log = tl.trace(tg_fn,       inputs, backend='tinygrad')
log = tl.trace(paddle_model, x,     backend='paddle')
log = tl.trace(tf_model,    x,      backend='tf')

PyTorch remains the full-feature backend. Preview backends are pinned and documented in docs/.

Gallery

TorchLens visualizes any architecture -- no matter how exotic. Below is a sample across families. The full menagerie has 650+ graphs across 44 architecture families.

Classic CNN + Vision Transformer

GoogLeNet (inception + buffer edges)	Stable Diffusion (U-Net denoiser)	CLIP (vision + language towers)

State-Space + Recurrence

Mamba (selective SSM)	Recurrent Gemma (linear recurrence)	Whisper (audio encoder-decoder)

Mixture-of-Experts + Generative

Mixtral (sparse MoE)	Hierarchical VAE	Perceiver

Graph Networks + Exotic

DimeNet (molecular GNN)	CORnet-S (visual cortex, unrolled)	LLaMA (decoder-only LLM)

Reinforcement Learning + Quantum ML + Scale

Decision Transformer (offline RL)	Quantum ML circuit	3,000-node graph (SFDP layout)

Compatibility

Before filing a bug for a model-specific failure, run the runtime compatibility report:

compat = tl.compat.report(model, x)
print(compat.to_markdown())

tl.compat.report inspects the model wrapper, modules, parameter sharing, input tensors, CUDA visibility, and common framework markers, then reports each row as pass, known_broken, scope, or not_tested.

TorchLens is not compatible with torch.compile'd models, TorchScript, or torch.export -- the forward pass does not run as ordinary Python, so the wrappers cannot intercept ops. It also has specific behaviors around FSDP, sparse tensors, meta tensors, quantization, and torch.func.vmap.

See LIMITATIONS.md for the full matrix: what fails, what works, and the recommended workaround for each context.

Tutorials and Docs

Resource	Description
torchlens_in_10_minutes.ipynb	Core workflow: trace, index, visualize
facets_tutorial.ipynb	Attention heads, LSTM facets, patching
backward_tutorial.ipynb	Gradient capture and backward visualization
training_tutorial.ipynb	Training with captured activations
huggingface_tutorial.ipynb	HuggingFace transformer models
fastlog_tutorial.ipynb	High-throughput sparse recording
docs/intervention_api.md	Full selector and helper reference
docs/backward.md	Backward capture details and limitations
docs/facets.md	Facets, patching, and SDPA reconstruction
docs/performance.md	Speed knobs and benchmark numbers

Security

Portable bundles contain a pickle file in metadata.pkl. Only load bundles from trusted sources. Loading an untrusted bundle with tl.load() can execute arbitrary code.

Other Packages You Should Check Out

TorchLens focuses on activation extraction, graph visualization, and intervention and intentionally omits model loading, stimulus management, and analysis pipelines. These packages cover that ground well:

• Cerbrec: interactive visualization and debugging for deep neural networks (uses TorchLens under the hood for PyTorch graph extraction)
• ThingsVision: model loading, stimulus management, and representational analysis for vision models
• Net2Brain: end-to-end pipeline for comparing DNN representations to neural data
• surgeon-pytorch: lightweight activation extraction with training-loss hooks
• deepdive: model loading and benchmarking across many model families
• torchvision feature_extraction: fast activation extraction for models with static computational graphs
• rsatoolbox: representational similarity analysis for DNN activations and brain data

Acknowledgments

The development of TorchLens benefitted greatly from discussions with Nikolaus Kriegeskorte, George Alvarez, Alfredo Canziani, Tal Golan, and the Visual Inference Lab at Columbia University. Thank you to Kale Kundert for helpful discussion and code contributions enabling PyTorch Lightning compatibility. Network visualizations are generated with Graphviz. Logo created by Nikolaus Kriegeskorte.

Citing TorchLens

To cite TorchLens, please cite this paper:

Taylor, J., Kriegeskorte, N. Extracting and visualizing hidden activations and computational graphs of PyTorch models with TorchLens. Sci Rep 13, 14375 (2023). https://doi.org/10.1038/s41598-023-40807-0

If you find TorchLens useful, a star on this repo is appreciated.

Contact

TorchLens is in active development. Questions, bug reports, and suggestions are welcome via email, Twitter, the issues page, or the discussion board.