© 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 EvalsExportsinterp-engineNEWAPI 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. 4374
    Prev
    Next
    INDEX
    Explanations

    mentions of political figures, particularly Silvio Berlusconi

    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
    447
    +0.13
    0.6%
    481
    +0.12
    0.5%
    684
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.13
    0.07
    659
    +0.12
    0.07
    265
    +0.11
    0.05
    Negative Logits
     *.
    -0.77
     paranormal
    -0.69
    Enlarge
    -0.65
     Thumbnails
    -0.65
    Materials
    -0.65
    \">
    -0.64
     Triangle
    -0.64
    Ô
    -0.63
     [&
    -0.61
    flies
    -0.60
    POSITIVE LOGITS
    addafi
    0.92
    uterte
    0.87
     Obama
    0.85
    Obama
    0.84
    anyahu
    0.82
    ovich
    0.81
     Jindal
    0.81
    iannopoulos
    0.80
    lopp
    0.80
     Hollande
    0.80
    Activations Density 0.201%

    No Known Activations