© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 10061
    Prev
    Next
    INDEX
    Explanations

    contractions of the words "would" and "not"

    oai_token-act-pair · gpt-3.5-turbo

    negations expressed through the word "wouldn't"

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     present
    -0.56
    ipes
    -0.55
     beginning
    -0.55
     Appendix
    -0.55
     med
    -0.54
    mine
    -0.54
     medium
    -0.53
     various
    -0.53
     preparation
    -0.51
     See
    -0.51
    POSITIVE LOGITS
     wouldn
    3.07
     Wouldn
    2.12
     couldn
    2.06
     shouldn
    1.95
     hadn
    1.85
     wasn
    1.84
     didn
    1.78
     weren
    1.76
     would
    1.68
    would
    1.68
    Activations Density 0.010%

    No Known Activations