© 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 24576
    9. 8-RES_FS24576-JB
    11. 20047
    Prev
    Next
    INDEX
    Explanations

    references to specific political figures

    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
    99
    +0.16
    0.7%
    288
    +0.11
    0.5%
    350
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    99
    +0.16
    0.02
    224
    +0.11
    0.02
    350
    +0.11
    0.01
    Negative Logits
    ournal
    -0.80
    ++++++++++++++++
    -0.77
    raints
    -0.73
    velength
    -0.70
    ++++
    -0.69
    LESS
    -0.69
    pee
    -0.68
    Constructed
    -0.68
    ibaba
    -0.67
    jer
    -0.64
    POSITIVE LOGITS
     Gingrich
    1.48
     Newt
    0.82
    ervative
    0.78
    enstein
    0.75
    mann
    0.74
     bund
    0.73
    cong
    0.72
     Rove
    0.72
     Cain
    0.68
     Giuliani
    0.68
    Activations Density 0.024%

    No Known Activations