© 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. 2439
    Prev
    Next
    INDEX
    Explanations

    references to historical events, locations, or artifacts

    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
    288
    +0.14
    0.6%
    57
    +0.13
    0.6%
    713
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    713
    +0.14
    0.02
    288
    +0.13
    0.02
    57
    +0.11
    0.02
    Negative Logits
    ertodd
    -1.00
    lain
    -0.96
    nery
    -0.88
    geon
    -0.88
    DAQ
    -0.79
    Flow
    -0.77
    ccoli
    -0.74
    igans
    -0.73
    loo
    -0.73
    gur
    -0.72
    POSITIVE LOGITS
     significance
    0.99
     precedent
    0.98
     preservation
    0.96
     relics
    0.93
    orical
    0.90
     revision
    0.90
     accur
    0.87
    orically
    0.87
     landmarks
    0.87
     preced
    0.86
    Activations Density 0.035%

    No Known Activations