© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 8094
    Prev
    Next
    INDEX
    Explanations

    punctuation marks, specifically commas

    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
    447
    +0.19
    0.9%
    480
    +0.18
    0.8%
    231
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    231
    +0.19
    0.05
    29
    +0.18
    0.05
    696
    +0.10
    0.05
    Negative Logits
     drawer
    -0.71
     spont
    -0.69
     gag
    -0.66
     rigs
    -0.62
     misunder
    -0.62
    "},"
    -0.60
     appe
    -0.59
     thous
    -0.57
    ussy
    -0.57
     comprom
    -0.57
    POSITIVE LOGITS
    actionDate
    0.92
     ][
    0.78
    ojure
    0.70
    align
    0.70
    taboola
    0.69
    ĸļ
    0.69
    then
    0.68
    à¼
    0.68
    ,,,,,,,,
    0.64
    ::::::::
    0.63
    Activations Density 0.060%

    No Known Activations