© 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. 14297
    Prev
    Next
    INDEX
    Explanations

    phrases related to work activities and commitments

    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
    352
    +0.11
    0.5%
    490
    +0.09
    0.4%
    166
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    718
    +0.11
    0.04
    331
    +0.09
    0.03
    513
    +0.09
    0.03
    Negative Logits
    mentioned
    -0.68
    uctions
    -0.63
    PLA
    -0.62
    rises
    -0.59
    mares
    -0.58
    notations
    -0.58
     trusts
    -0.57
    endas
    -0.57
    çīĪ
    -0.56
    utsu
    -0.56
    POSITIVE LOGITS
    renheit
    0.71
    cule
    0.70
     unto
    0.68
    dimensional
    0.68
     PIT
    0.63
     understatement
    0.63
     annoyance
    0.61
     akin
    0.61
    parable
    0.60
     damn
    0.60
    Activations Density 0.208%

    No Known Activations