© 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 · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 13158
    Prev
    Next
    INDEX
    Explanations

    phrases or sentences starting with the word "We."

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    559
    +0.13
    0.6%
    354
    +0.12
    0.6%
    167
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    354
    +0.13
    0.02
    559
    +0.12
    0.01
    234
    +0.10
    0.01
    Negative Logits
    BACK
    -0.73
    nesota
    -0.72
     Eleven
    -0.71
    liest
    -0.68
     Clown
    -0.63
     UD
    -0.61
    aults
    -0.61
     partName
    -0.61
    rising
    -0.59
    past
    -0.59
    POSITIVE LOGITS
    bsite
    0.89
    athered
    0.87
    ibo
    0.85
    selves
    0.85
    eding
    0.83
    bley
    0.82
    're
    0.82
     âĿ
    0.80
    onut
    0.79
    ird
    0.79
    Activations Density 0.016%

    No Known Activations