© 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. 14448
    Prev
    Next
    INDEX
    Explanations

    keywords related to programming logic, such as syntax elements and coding conventions

    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
    596
    +0.10
    0.5%
    485
    +0.09
    0.4%
    8
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    8
    +0.10
    0.01
    248
    +0.09
    0.01
    485
    +0.09
    0.01
    Negative Logits
    natureconservancy
    -0.64
    istle
    -0.57
     cle
    -0.56
    levard
    -0.53
     straw
    -0.52
     possessed
    -0.52
     Mü
    -0.52
    detail
    -0.52
    vd
    -0.51
     biomedical
    -0.51
    POSITIVE LOGITS
    vier
    0.86
    where
    0.78
    wise
    0.75
    leans
    0.75
    ption
    0.71
    phan
    0.71
    if
    0.69
    lando
    0.67
    ional
    0.67
    hedon
    0.65
    Activations Density 0.019%

    No Known Activations