© 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 EvalsExportsAPI 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. 6258
    Prev
    Next
    INDEX
    Explanations

    words related to the creative process, storytelling, design, and specific details related to film and television productions

    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
    163
    +0.11
    0.5%
    701
    +0.10
    0.5%
    116
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    701
    +0.11
    0.02
    232
    +0.10
    0.02
    299
    +0.10
    0.02
    Negative Logits
     segregation
    -0.61
     tampering
    -0.60
     Mub
    -0.60
     consecut
    -0.58
     gestation
    -0.57
     landfall
    -0.56
     housing
    -0.56
     conduc
    -0.56
     circulation
    -0.55
     correctness
    -0.55
    POSITIVE LOGITS
    ilus
    0.75
    ogyn
    0.73
    iotic
    0.70
    esthetic
    0.69
    elist
    0.68
     Truly
    0.66
    brand
    0.66
    ãĤ¼
    0.66
    obic
    0.66
    tre
    0.65
    Activations Density 0.048%

    No Known Activations