© 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. 15066
    Prev
    Next
    INDEX
    Explanations

    statements or opinions being challenged or criticized

    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
    509
    +0.13
    0.6%
    446
    +0.12
    0.6%
    745
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    509
    +0.13
    0.05
    540
    +0.12
    0.06
    386
    +0.10
    0.05
    Negative Logits
    regor
    -0.71
     oneself
    -0.69
    glomer
    -0.66
    ividually
    -0.65
    Jew
    -0.65
     signific
    -0.63
    oke
    -0.63
    byter
    -0.62
    hma
    -0.61
    rafted
    -0.59
    POSITIVE LOGITS
     regarding
    0.78
    lessness
    0.77
    fulness
    0.73
    steps
    0.68
     spree
    0.67
    naires
    0.66
    less
    0.65
     amnesty
    0.65
    making
    0.64
     foray
    0.64
    Activations Density 0.217%

    No Known Activations