© 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. 22332
    Prev
    Next
    INDEX
    Explanations

    certain repeated phrases or terms, specifically focusing on the word "the."

    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
    393
    +0.19
    0.9%
    362
    +0.15
    0.7%
    608
    +0.14
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.19
    0.11
    608
    +0.15
    0.10
    385
    +0.14
    0.09
    Negative Logits
     preceded
    -0.71
    opsis
    -0.71
    thood
    -0.71
     namely
    -0.70
    tackle
    -0.70
    scape
    -0.69
    VERTISEMENT
    -0.66
     compared
    -0.65
    agree
    -0.65
    verage
    -0.65
    POSITIVE LOGITS
     feds
    1.02
     highest
    0.99
     aforementioned
    0.98
     Department
    0.95
     Dalai
    0.94
     likes
    0.93
     respective
    0.93
     Clintons
    0.93
     mayor
    0.93
     legislature
    0.92
    Activations Density 0.382%

    No Known Activations