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

    Neuronpedia

    Jacobian LensNEW
    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. 6496
    Prev
    Next
    INDEX
    Explanations

    references to specific historical figures, particularly Hitler and Stalin

    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
    +0.17
    0.8%
    288
    +0.14
    0.7%
    218
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    346
    +0.17
    0.02
    288
    +0.14
    0.02
    57
    +0.13
    0.02
    Negative Logits
    Dub
    -0.76
    player
    -0.72
    ratom
    -0.70
    Interstitial
    -0.69
    TOR
    -0.69
    pole
    -0.69
    forward
    -0.65
    olulu
    -0.64
    Asset
    -0.64
    notes
    -0.64
    POSITIVE LOGITS
     Hitler
    1.10
    ocaust
    0.94
     Osw
    0.91
     Youth
    0.85
     Adolf
    0.82
     dictator
    0.82
    wald
    0.81
    chwitz
    0.81
    enthal
    0.81
    olini
    0.81
    Activations Density 0.036%

    No Known Activations