© 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 EvalsExportsinterp-engineNEWAPI 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. 13655
    Prev
    Next
    INDEX
    Explanations

    instances of the word "brush" in the text

    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
    314
    +0.12
    0.5%
    285
    +0.12
    0.5%
    289
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    212
    +0.12
    0.01
    349
    +0.12
    0.01
    285
    +0.10
    0.01
    Negative Logits
     Founding
    -0.70
     Defenders
    -0.69
     regress
    -0.69
    reme
    -0.68
    infeld
    -0.67
    uthor
    -0.66
    oppable
    -0.64
     Mandela
    -0.64
    CVE
    -0.63
    rophe
    -0.63
    POSITIVE LOGITS
    brush
    1.09
    stro
    0.93
    stroke
    0.87
     strokes
    0.85
     brush
    0.84
     applic
    0.83
    sonian
    0.83
    tip
    0.83
     brist
    0.81
    lessly
    0.80
    Activations Density 0.020%

    No Known Activations