© 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. 3092
    Prev
    Next
    INDEX
    Explanations

    questions and statements related to reasons and explanations

    oai_token-act-pair · gpt-3.5-turbo
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    373
    +0.29
    1.4%
    273
    +0.09
    0.4%
    426
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    273
    +0.29
    0.07
    267
    +0.09
    0.06
    50
    +0.09
    0.05
    Negative Logits
     ..."
    -0.97
     ..
    -0.66
     wi
    -0.65
     ...
    -0.65
     ....
    -0.64
    .","
    -0.64
     "-
    -0.63
     etc
    -0.61
    "],"
    -0.60
     orbs
    -0.60
    POSITIVE LOGITS
    odore
    1.06
    resa
    1.05
    xiety
    1.01
     Facts
    0.90
    withstanding
    0.88
    alyst
    0.83
    hibit
    0.83
    aminer
    0.82
    swers
    0.80
    undrum
    0.77
    Activations Density 0.507%

    No Known Activations