© 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. 13067
    Prev
    Next
    INDEX
    Explanations

    specific tags or classes in a structured document, likely related to HTML or formatting

    oai_token-act-pair · gpt-3.5-turbo

    instances of the "cd" product designation in a document

    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
    FIN
    -0.71
     Finland
    -0.66
     Arena
    -0.65
     tight
    -0.64
     Pride
    -0.63
     soy
    -0.63
     Richardson
    -0.63
     Uzbek
    -0.61
     Gron
    -0.61
    ön
    -0.61
    POSITIVE LOGITS
    cd
    3.56
     cd
    3.20
    cf
    1.43
    CDC
    1.43
    CD
    1.35
    cb
    1.30
     cp
    1.29
    bd
    1.27
    cp
    1.24
    fd
    1.23
    Activations Density 0.029%

    No Known Activations