© 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. 24253
    Prev
    Next
    INDEX
    Explanations

    instances of the word "democratic."

    oai_token-act-pair · gpt-4o-miniTriggered 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
    316
    +0.12
    0.5%
    167
    +0.11
    0.5%
    389
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    389
    +0.12
    0.01
    137
    +0.11
    0.01
    316
    +0.10
    0.01
    Negative Logits
    WAYS
    -0.73
     Elves
    -0.72
     supper
    -0.67
    FORE
    -0.66
    tis
    -0.65
    nce
    -0.63
    glers
    -0.62
    Dub
    -0.62
    loo
    -0.61
     Indigo
    -0.61
    POSITIVE LOGITS
    agogue
    1.25
    agog
    1.11
    aterial
    1.10
    ilit
    1.06
    olition
    1.06
    ographically
    1.05
    otions
    1.04
    igration
    1.04
    otion
    1.03
    ploy
    1.03
    Activations Density 0.008%

    No Known Activations