© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 1408
    Prev
    Next
    INDEX
    Explanations

    proper nouns, especially related to political figures, locations, or organizations

    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
    481
    +0.30
    1.4%
    459
    +0.12
    0.6%
    324
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.30
    0.09
    481
    +0.12
    0.08
    225
    +0.12
    0.07
    Negative Logits
    pires
    -0.72
    ãĥĦ
    -0.57
     sleeps
    -0.56
    ãĥ¯ãĥ³
    -0.54
     ceases
    -0.54
     guiIcon
    -0.53
    ãĥīãĥ©
    -0.52
    guyen
    -0.52
    Ö¼
    -0.52
    ª
    -0.50
    POSITIVE LOGITS
     respectively
    1.36
     apiece
    1.24
     themselves
    0.85
     whereas
    0.83
     nowadays
    0.73
    their
    0.73
    .
    0.72
     anyways
    0.70
     because
    0.70
    *.
    0.70
    Activations Density 0.840%

    No Known Activations