© 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. 22374
    Prev
    Next
    INDEX
    Explanations

    references to various types of creatures

    oai_token-act-pair · gpt-4o-miniTriggered 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
    575
    +0.12
    0.6%
    160
    +0.10
    0.5%
    414
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    575
    +0.12
    0.02
    548
    +0.10
    0.01
    762
    +0.10
    0.02
    Negative Logits
    olitan
    -0.73
     Boys
    -0.73
    Tx
    -0.72
     Gin
    -0.69
     caut
    -0.68
    dL
    -0.64
    keye
    -0.62
    ierra
    -0.62
    road
    -0.61
    outer
    -0.61
    POSITIVE LOGITS
    arium
    1.01
     creatures
    0.93
     inhab
    0.90
     vanquished
    0.90
     resembling
    0.89
     creature
    0.88
    folk
    0.84
     beings
    0.83
     incapable
    0.77
     sightings
    0.76
    Activations Density 0.040%

    No Known Activations