© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 24529
    Prev
    Next
    INDEX
    Explanations

    quantifiable statistics or numerical data

    oai_token-act-pair · gpt-4o-miniTriggered 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
    742
    +0.11
    0.5%
    455
    +0.10
    0.5%
    405
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    742
    +0.11
    0.04
    310
    +0.10
    0.04
    405
    +0.10
    0.04
    Negative Logits
    venge
    -0.70
    Redd
    -0.67
    vised
    -0.67
    eworld
    -0.64
    ãĤ´ãĥ³
    -0.62
    laden
    -0.60
    incarn
    -0.59
    kefeller
    -0.59
    jri
    -0.59
     conquering
    -0.58
    POSITIVE LOGITS
     respondents
    0.95
     percent
    0.83
     Australians
    0.79
     Britons
    0.77
     Americans
    0.76
     Canadians
    0.70
    percent
    0.69
     Percent
    0.68
     of
    0.67
     disapprove
    0.66
    Activations Density 0.084%

    No Known Activations