© 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 12288
    9. 8-RES_FS12288-JB
    11. 10799
    Prev
    Next
    INDEX
    Explanations

    personal disbelief or surprise

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    373
    +0.11
    0.5%
    92
    +0.10
    0.5%
    267
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    510
    +0.11
    0.03
    750
    +0.10
    0.03
    610
    +0.09
    0.02
    Negative Logits
    edom
    -0.70
    catentry
    -0.69
    Footnote
    -0.68
     sidx
    -0.64
    esthetic
    -0.63
    bis
    -0.63
    onding
    -0.63
    cially
    -0.62
    atten
    -0.62
    deep
    -0.61
    POSITIVE LOGITS
    enance
    0.79
     coincidence
    0.73
     Saban
    0.69
     THAT
    0.66
    biz
    0.65
    IAN
    0.65
    !--
    0.65
     how
    0.61
     Frog
    0.60
    !!!!!
    0.60
    Activations Density 0.162%

    No Known Activations