© 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. 22731
    Prev
    Next
    INDEX
    Explanations

    mentions of prototypes and prototype-related concepts in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered 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
    288
    +0.13
    0.6%
    541
    +0.12
    0.6%
    134
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    541
    +0.13
    0.02
    288
    +0.12
    0.02
    134
    +0.11
    0.01
    Negative Logits
    leans
    -0.89
    die
    -0.82
    olulu
    -0.79
    tics
    -0.79
    iane
    -0.78
    andra
    -0.77
    iris
    -0.76
    wind
    -0.75
    relative
    -0.74
    lez
    -0.73
    POSITIVE LOGITS
     prototype
    1.02
     prototypes
    0.93
    smanship
    0.87
     sketches
    0.82
     mock
    0.82
    hesis
    0.78
     kit
    0.78
     kits
    0.78
     sketch
    0.78
     Prototype
    0.76
    Activations Density 0.035%

    No Known Activations