© 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. 5084
    Prev
    Next
    INDEX
    Explanations

    cooking instructions and ingredients for a recipe

    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
    589
    +0.11
    0.5%
    238
    +0.10
    0.5%
    95
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    545
    +0.11
    0.02
    678
    +0.10
    0.02
    238
    +0.09
    0.02
    Negative Logits
     envy
    -0.68
     Nug
    -0.66
    ticket
    -0.66
     purchaser
    -0.63
     sponsor
    -0.61
     recru
    -0.61
    ©¶æ
    -0.61
     grou
    -0.60
     redistributed
    -0.59
     editor
    -0.59
    POSITIVE LOGITS
    Ingredients
    0.82
    arella
    0.81
    utes
    0.80
    Slow
    0.79
    cation
    0.75
    poons
    0.74
     intervals
    0.72
    illi
    0.71
     Slowly
    0.69
    uted
    0.69
    Activations Density 0.038%

    No Known Activations