© 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. 17005
    Prev
    Next
    INDEX
    Explanations

    words related to coding or technical terms

    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
    659
    +0.17
    0.8%
    752
    +0.11
    0.5%
    578
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.17
    0.04
    752
    +0.11
    0.04
    251
    +0.10
    0.03
    Negative Logits
     GOODMAN
    -0.90
    ãĥĹ
    -0.69
     Tinder
    -0.65
     Chero
    -0.65
     Sacrament
    -0.63
     Yamaha
    -0.62
    tenance
    -0.61
     Dish
    -0.60
     disabling
    -0.60
     Bain
    -0.60
    POSITIVE LOGITS
    rosso
    1.32
    ologne
    1.32
    rescent
    1.28
    ordon
    1.23
    auldron
    1.23
    elsius
    1.21
    heddar
    1.20
    aucas
    1.19
    actus
    1.19
    atered
    1.14
    Activations Density 0.025%

    No Known Activations