© 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. 20450
    Prev
    Next
    INDEX
    Explanations

    keywords related to digital art and online gambling

    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
    373
    +0.20
    1.0%
    674
    +0.10
    0.5%
    598
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.20
    0.06
    530
    +0.10
    0.03
    373
    +0.09
    -0.04
    Negative Logits
    hran
    -0.79
     disapp
    -0.72
    Ö¼
    -0.62
    instead
    -0.62
    essage
    -0.61
    some
    -0.61
     ."
    -0.61
     destro
    -0.60
    .ãĢį
    -0.60
    advertising
    -0.60
    POSITIVE LOGITS
     Profile
    1.03
     Version
    0.87
     Definitions
    0.85
     Description
    0.85
     Variant
    0.84
     Discussion
    0.83
     References
    0.83
    %:
    0.83
     Translation
    0.82
     Miscellaneous
    0.80
    Activations Density 0.435%

    No Known Activations