© 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. 20
    Prev
    Next
    INDEX
    Explanations

    mentions of luxury car brands like Mercedes, as well as specific models and features of Mercedes vehicles

    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
    285
    +0.10
    0.5%
    326
    +0.10
    0.5%
    538
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.10
    0.01
    538
    +0.10
    0.01
    343
    +0.10
    0.01
    Negative Logits
    yip
    -0.96
    umbn
    -0.80
    aji
    -0.80
    ricular
    -0.78
    agnar
    -0.77
    omial
    -0.75
    tery
    -0.73
    ally
    -0.72
    ochond
    -0.72
    tin
    -0.72
    POSITIVE LOGITS
     Benz
    1.43
    Benz
    1.23
    cedes
    1.08
     Motors
    0.75
    bach
    0.71
    manship
    0.69
     Sachs
    0.67
     Mercedes
    0.66
     Motor
    0.66
     Motorsport
    0.65
    Activations Density 0.020%

    No Known Activations