© 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 24576
    9. 8-RES_FS24576-JB
    11. 2042
    Prev
    Next
    INDEX
    Explanations

    references to scientific research articles and their authors

    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
    46
    +0.09
    0.4%
    106
    +0.09
    0.4%
    235
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    46
    +0.09
    0.05
    106
    +0.09
    0.05
    539
    +0.09
    0.04
    Negative Logits
    estate
    -0.79
    iquette
    -0.74
    deal
    -0.73
    netflix
    -0.71
    wives
    -0.70
     cess
    -0.70
    ince
    -0.68
    compl
    -0.68
     condos
    -0.68
     stale
    -0.67
    POSITIVE LOGITS
     researchers
    1.30
     researcher
    1.29
     doctoral
    1.21
     PhD
    1.20
     Neuroscience
    1.18
     Researchers
    1.17
     scientists
    1.16
     neuroscience
    1.15
     scientist
    1.14
     Professor
    1.11
    Activations Density 0.344%

    No Known Activations