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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 12288
    9. 8-RES_FS12288-JB
    11. 10289
    Prev
    Next
    INDEX
    Explanations

    names starting with "Mrs" or "Mrs."

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    123
    +0.11
    0.5%
    430
    +0.10
    0.5%
    642
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    430
    +0.11
    0.03
    232
    +0.10
    0.02
    754
    +0.09
    0.02
    Negative Logits
    umbn
    -0.90
    undo
    -0.88
    osta
    -0.79
    asper
    -0.75
    amaru
    -0.75
    anwhile
    -0.74
    adiator
    -0.74
    ioch
    -0.74
    anamo
    -0.70
    ographed
    -0.69
    POSITIVE LOGITS
     Thatcher
    0.94
     Mae
    0.92
     Teresa
    0.86
     herself
    0.85
     Theresa
    0.83
     Merkel
    0.83
     Clinton
    0.82
     Louise
    0.82
     DeVos
    0.81
     Claus
    0.81
    Activations Density 0.042%

    No Known Activations