© 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. 15768
    Prev
    Next
    INDEX
    Explanations

    repetition within sentences

    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
    191
    +0.15
    0.7%
    341
    +0.11
    0.5%
    543
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    227
    +0.15
    0.02
    48
    +0.11
    0.02
    341
    +0.09
    0.02
    Negative Logits
    tec
    -0.79
    nex
    -0.77
    erb
    -0.73
    estial
    -0.71
    atom
    -0.69
     Nun
    -0.69
    cht
    -0.67
    anan
    -0.66
    tc
    -0.66
    oons
    -0.66
    POSITIVE LOGITS
     importing
    1.30
     transferring
    1.28
     owning
    1.28
     applying
    1.24
     registering
    1.23
     exercising
    1.22
     treating
    1.22
     putting
    1.22
     creating
    1.22
     bringing
    1.22
    Activations Density 0.208%

    No Known Activations