sae-lens

library6.44.3pypypi✓ verified 84d ago

SAE Lens is a library for training, loading, and analyzing sparse autoencoders (SAEs) on transformer language models. Current version is 6.43.0, with frequent releases (multiple versions per month).

ai-ml llm-agents

pip install sae-lens

github ↗package ↗homepage ↗

INSTALL

IMPORT

SIG · SAE-LENS

sae-lens

ai-mlpythonv6.44.3

Install

—

Import

—

Disk

—

Pass rate

0/ 10

Env Coverage0 / 10

glibc

3.9–3.13

musl

3.9–3.13

Install & Compatibility

Where this runs

tested against v? · pip install

Install × environment matrix

Each cell = how many times install + import succeeded across repeated harness runs. Partial = flaky.
glibc = Debian/Ubuntu slim · musl = Alpine Linux

musl

py 3.10–3.95 runs

build_error

glibc

py 3.10–3.95 runs

timeout

Code

Verified usage

Verified import paths — ran on the pinned version, not inferred.

SAE

✓ from sae_lens import SAE

Standard import for loading a pretrained SAE.

SAEConfig

✓ from sae_lens.config import SAEConfig

✗ from sae_lens import SAEConfig

SAEConfig is in sae_lens.config, not top-level.

HookedSAETransformer

✓ from sae_lens import HookedSAETransformer

Wraps a HookedTransformer to cache activations.

Load a pretrained SAE and compute feature activations for a prompt.

from sae_lens import SAE from transformer_lens import HookedTransformer model = HookedTransformer.from_pretrained("gpt2-small", device="cpu") sae, cfg_dict, sparsity = SAE.from_pretrained(release="gpt2-small-res-jb", sae_id="blocks.0.hook_resid_pre", device="cpu") sae.to("cpu") # Example: get SAE feature activations for a prompt prompt = "Hello, world!" _, cache = model.run_with_cache(prompt, names_filter=[sae.cfg.hook_name]) act = cache[sae.cfg.hook_name] sae_acts = sae.encode(act) print(sae_acts.shape)

Debug

Known issues

breakingIn v6.x, the `SAE.from_pretrained` signature changed: the `release` argument is now the first positional argument and required. Old code using `SAE.from_pretrained(sae_id=...)` without `release` will break.

fix

Update to: SAE.from_pretrained(release=..., sae_id=...)

affects: <6.0

gotchaSAE expects activations on the same device as the SAE itself. Cross-device (e.g., model on GPU, SAE on CPU) can cause silent errors or crashes.

fix

Ensure model and SAE are on the same device (use sae.to(device) and model.to(device)).

affects: all

deprecatedThe `cache` parameter in `SAE.encode` is deprecated and will be removed in a future version. Use `HookedSAETransformer` or `model.run_with_cache` explicitly.

fix

Switch to using `model.run_with_cache` and pass the activations directly to `sae.encode`.

affects: >=6.30.0

Errors

Common errors & fixes

ModuleNotFoundError: No module named 'sae_lens'

Package not installed.

fix

Run `pip install sae-lens`.

AssertionError: Expected activation shape (batch, seq_len, d_model) but got ...

Activations passed to SAE.encode have incorrect shape or are on wrong device.

fix

Check that activations are a 3D tensor on the same device as the SAE. Use `sae.encode(act)` where `act` is shape (batch, seq_len, d_model).

ValueError: Unknown release: ...

Provided release name does not exist in the SAE registry.

fix

Use a valid release name from `sae_lens.known_releases()` or check the docs.

Upgrade

Version history

6.44.3latest on PyPI · released Jun 16, 2026

Audit

Dependencies

torchrequiredCore dependency for tensor operations and model loading

transformer-lensrequiredUsed for model hooks and activation caching

datasetsrequiredFor loading activation datasets

wandboptionalOptional for experiment logging

Agent activity

27 hits · last 30 days

node

OpenAI (training)

Resources

githubhttps://github.com/decoderesearch/SAELens ↗

packagehttps://pypi.org/project/sae-lens/ ↗

homepagehttps://decoderesearch.github.io/SAELens ↗