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

    Neuronpedia

    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 12288
    9. 8-RES_FS12288-JB
    11. 10609
    Prev
    Next
    INDEX
    Explanations

    phrases conveying the concept of comparing or providing an understanding of something through examples or explanations

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    413
    +0.11
    0.5%
    62
    +0.11
    0.5%
    513
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    413
    +0.11
    0.04
    513
    +0.11
    0.04
    568
    +0.10
    0.03
    Negative Logits
    iki
    -0.70
    die
    -0.65
    akia
    -0.63
    yne
    -0.60
     Cosmos
    -0.56
     abolished
    -0.56
    vertisement
    -0.55
     Budapest
    -0.55
    jl
    -0.54
     Meyer
    -0.54
    POSITIVE LOGITS
     how
    0.91
     whats
    0.89
     approximation
    0.84
     whereabouts
    0.81
     guesses
    0.78
     rundown
    0.78
     thereof
    0.77
     why
    0.76
     estimate
    0.76
     impression
    0.75
    Activations Density 0.254%

    No Known Activations