© 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. 5166
    Prev
    Next
    INDEX
    Explanations

    detailed descriptions of situations or events

    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
    640
    +0.35
    1.7%
    480
    +0.17
    0.8%
    314
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.35
    0.06
    328
    +0.17
    0.03
    464
    +0.12
    0.03
    Negative Logits
    hement
    -0.80
     diving
    -0.73
     steroids
    -0.69
     dissu
    -0.68
     Scrib
    -0.66
     steering
    -0.66
     charm
    -0.63
     tackling
    -0.63
     outwe
    -0.61
     answering
    -0.61
    POSITIVE LOGITS
    ³³³
    1.18
    ³³³³³³³³
    1.12
    ³³³³³³³³³³³³³³³³
    1.12
    ³³³³
    1.11
    ccording
    0.94
    pmwiki
    0.88
    ³³
    0.87
     largeDownload
    0.85
    PRESS
    0.84
    ente
    0.83
    Activations Density 0.103%

    No Known Activations