© 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. 5420
    Prev
    Next
    INDEX
    Explanations

    legal and bureaucratic terms associated with federal funding and government projects

    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
    +0.31
    1.5%
    481
    +0.17
    0.8%
    447
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.31
    0.17
    442
    +0.17
    0.15
    107
    +0.12
    0.13
    Negative Logits
     colours
    -0.88
     humour
    -0.87
     endeavour
    -0.82
     outings
    -0.81
     thrill
    -0.80
     regener
    -0.77
     util
    -0.77
     organising
    -0.77
     enqu
    -0.77
     worsh
    -0.77
    POSITIVE LOGITS
    According
    1.54
    Advertisement
    1.51
    Specifically
    1.46
    Asked
    1.43
    However
    1.43
    Related
    1.38
    Article
    1.38
    RELATED
    1.37
    Documents
    1.36
    Instead
    1.36
    Activations Density 0.520%

    No Known Activations