© 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. 20878
    Prev
    Next
    INDEX
    Explanations

    phrases involving actions, physical objects, and activities related to writing and creativity

    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
    480
    +0.39
    1.9%
    373
    +0.15
    0.7%
    678
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    678
    +0.39
    0.05
    526
    +0.15
    -0.01
    555
    +0.10
    0.04
    Negative Logits
     historically
    -0.66
     domestically
    -0.63
     nowadays
    -0.62
    uably
    -0.62
     geographically
    -0.59
     politically
    -0.59
    ossier
    -0.58
     Historically
    -0.56
     Polit
    -0.56
     tainted
    -0.55
    POSITIVE LOGITS
    FIR
    0.65
     gently
    0.62
     hurried
    0.61
    gently
    0.60
    zzle
    0.59
     breakfast
    0.59
    cellaneous
    0.59
     tray
    0.59
    grim
    0.58
     downstairs
    0.56
    Activations Density 0.572%

    No Known Activations