© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 18858
    Prev
    Next
    INDEX
    Explanations

    numbers related to large quantities or high values

    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
    697
    +0.11
    0.5%
    355
    +0.10
    0.5%
    453
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    697
    +0.11
    0.03
    408
    +0.10
    0.04
    453
    +0.10
    0.03
    Negative Logits
    pring
    -0.84
    ringe
    -0.75
    Reviewed
    -0.75
    Dialog
    -0.73
    nav
    -0.71
    ynski
    -0.69
    rien
    -0.69
    actionDate
    -0.69
    abet
    -0.68
    imore
    -0.67
    POSITIVE LOGITS
     amounts
    1.26
     proportions
    1.24
     amount
    1.21
     quantities
    0.95
     heights
    0.91
     sums
    0.90
    amount
    0.88
     expense
    0.85
    ly
    0.85
     disproportion
    0.84
    Activations Density 0.124%

    No Known Activations