© 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. 2840
    Prev
    Next
    INDEX
    Explanations

    phrases related to erasing or covering up information, such as "whitewashed" and "whitewashing"

    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
    568
    +0.11
    0.5%
    444
    +0.10
    0.5%
    486
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.11
    -0.01
    103
    +0.10
    0.04
    232
    +0.10
    0.03
    Negative Logits
    )=(
    -0.77
     Inquis
    -0.73
     Jur
    -0.71
     Zur
    -0.63
     Trin
    -0.62
     epid
    -0.62
     respons
    -0.61
     Zan
    -0.61
     surg
    -0.61
     Gil
    -0.61
    POSITIVE LOGITS
    akers
    0.99
    ound
    0.98
    oop
    0.97
    oard
    0.94
    aper
    0.94
    ocking
    0.93
    ocket
    0.93
    ood
    0.92
    arter
    0.91
    acket
    0.91
    Activations Density 0.089%

    No Known Activations