TransformerLensOrg/TransformerLens

[Proposal] Add support for `cpu`, `meta`, and `disk` to TransformerBridge `device_map` about 5 hours ago

enhancement help wanted complexity-moderate TransformerBridge

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Additional Architecture Adapter tests about 11 hours ago

enhancement help wanted good first issue low-priority complexity-simple TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Migrate the dev/CI Jupyter stack to Notebook 7 about 17 hours ago

documentation enhancement help wanted minor demo complexity-moderate

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Centralize soft-cap logic behind a shared helper and a named "disabled" sentinel 1 day ago

enhancement help wanted good first issue complexity-simple

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add RWKV-7 "Goose" adapter (RWKV7ForCausalLM) 1 day ago

help wanted new-architecture complexity-high TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add Raven/Huginn depth-recurrent adapter (RavenForCausalLM) 1 day ago

help wanted new-architecture complexity-high TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add HRM-Text two-timescale recurrent adapter (HrmTextForCausalLM) 1 day ago

help wanted new-architecture complexity-high TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add Ouro LoopLM looped-depth adapter (OuroForCausalLM) 1 day ago

help wanted new-architecture complexity-high TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add Zamba2 shared-attention hybrid adapter (Zamba2ForCausalLM) 5 days ago

help wanted new-architecture complexity-moderate TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add RecurrentGemma (Griffin) adapter (RecurrentGemmaForCausalLM) 5 days ago

help wanted new-architecture complexity-high TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add Jamba attention+Mamba interleave adapter (JambaForCausalLM) 6 days ago

help wanted new-architecture complexity-moderate TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add BD3LM block-diffusion adapter (BD3LM) 6 days ago

help wanted new-architecture complexity-high TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add Arcee squared-ReLU sparse-MLP adapter (ArceeForCausalLM) 7 days ago

help wanted good first issue new-architecture complexity-moderate TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add Qwen2-MoE shared+routed expert adapter (Qwen2MoeForCausalLM) 7 days ago

help wanted good first issue new-architecture complexity-simple TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add Cohere2 / Command-A interleaved-attention adapter (Cohere2ForCausalLM) 7 days ago

help wanted good first issue new-architecture complexity-moderate TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add LFM2 dense short-convolution adapter (Lfm2ForCausalLM) 8 days ago

help wanted new-architecture complexity-moderate TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add LLaDA masked-diffusion adapter (LLaDAModelLM) 8 days ago

help wanted new-architecture complexity-high TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

Support T5Gemma2 (Gemma 3) Architecture 8 days ago

help wanted good first issue new-architecture complexity-moderate TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add HunYuan dense adapter (HunYuanDenseV1ForCausalLM) 12 days ago

help wanted good first issue new-architecture complexity-simple TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add Falcon-H1 parallel hybrid adapter (FalconH1ForCausalLM) 14 days ago

help wanted minor new-architecture complexity-high TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add BART encoder-decoder adapter (BartForConditionalGeneration) 14 days ago

help wanted new-architecture complexity-moderate TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

consequences of tokenizer divergence between TransformersLens and Huggingface Transformers 15 days ago

help wanted good first issue complexity-simple

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add Nemotron-H hybrid Mamba2-Transformer adapter 15 days ago

help wanted new-architecture low-priority complexity-high TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add DeepSeek-V2 architecture adapter (DeepseekV2ForCausalLM) 16 days ago

help wanted good first issue minor new-architecture complexity-simple TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add GLM-5 (DSA) adapter (GlmMoeDsaForCausalLM) 17 days ago

help wanted new-architecture low-priority complexity-high TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add T5Gemma encoder-decoder adapter (T5GemmaForConditionalGeneration) 19 days ago

help wanted good first issue new-architecture complexity-moderate TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Add GLM-4.5 MoE adapter (Glm4MoeForCausalLM) 20 days ago

help wanted good first issue new-architecture complexity-moderate TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Gemma4 Architecture Adapter 29 days ago

enhancement help wanted complexity-moderate TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Test Coverage] Fix SVD interpreter compatibility test fixtures 29 days ago

good first issue low-priority complexity-simple TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Bug Report] Fix dense component initialization for Pythia hook shapes 29 days ago

bug help wanted good first issue complexity-moderate TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Tutorial for "Real-Time Training Dynamics" (VSM Telemetry) about 1 month ago

enhancement help wanted good first issue minor complexity-moderate

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Test Coverage] Parameterize Cohere `logit_scale` fold test with a non-trivial scale about 1 month ago

good first issue low-priority complexity-simple TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Test Coverage] Generation compatibility for `TransformerBridge` models lacking native `generate()` about 1 month ago

help wanted good first issue low-priority complexity-moderate TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] CI HuggingFace Call Reduction about 2 months ago

enhancement good first issue low-priority complexity-moderate

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Direct Logit Attribution Tool about 2 months ago

enhancement help wanted good first issue minor complexity-moderate

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] External Architecture Registration about 2 months ago

documentation enhancement help wanted good first issue complexity-simple TransformerBridge

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Update Benchmarks & Verify Models to support Quantized models 2 months ago

enhancement help wanted minor complexity-moderate

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] MPS CI Runner Support 2 months ago

enhancement help wanted complexity-high

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Setup unit tests to cover model configurations 2 months ago

good first issue testing

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Have ActivationCache.get_full_resid_decomposition support passing in a vector/tensor to project onto 2 months ago

help wanted good first issue complexity-moderate

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Support Gemma 3 3 months ago

enhancement help wanted high-priority new-architecture complexity-moderate model-request

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

[Proposal] Support R1 Distills 5 months ago

enhancement help wanted high-priority complexity-simple model-request

TransformerLensOrg/TransformerLens

3.6K

A library for mechanistic interpretability of GPT-style language models

Python

Open Issues Need Help