© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 3280
    Prev
    Next
    INDEX
    Explanations

    references to upcoming events or announcements

    oai_token-act-pair · gpt-3.5-turbo
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    481
    +0.21
    1.0%
    447
    +0.10
    0.5%
    393
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    718
    +0.21
    0.06
    119
    +0.10
    0.04
    512
    +0.10
    0.04
    Negative Logits
    aughtered
    -0.66
    ograp
    -0.65
    Farm
    -0.63
    eneg
    -0.63
    Dom
    -0.62
    agg
    -0.62
    ucc
    -0.62
    Tweet
    -0.62
    Cor
    -0.62
    Sil
    -0.62
    POSITIVE LOGITS
    etheless
    0.79
     conclud
    0.67
     furthermore
    0.67
     doub
    0.66
     also
    0.66
     additionally
    0.65
     downside
    0.63
     refuel
    0.62
    again
    0.62
     ALSO
    0.61
    Activations Density 0.803%

    No Known Activations