© 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. 8832
    Prev
    Next
    INDEX
    Explanations

    prepositional phrases starting with "up."

    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
    123
    +0.14
    0.7%
    101
    +0.11
    0.5%
    166
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    123
    +0.14
    0.02
    250
    +0.11
    0.02
    725
    +0.10
    0.02
    Negative Logits
    hens
    -0.71
    fml
    -0.71
    imb
    -0.69
    hello
    -0.68
    iw
    -0.67
    edu
    -0.65
    Corp
    -0.65
     awa
    -0.64
    aed
    -0.62
     arri
    -0.61
    POSITIVE LOGITS
    illac
    0.81
     majority
    0.67
    Ĥª
    0.65
     Communities
    0.64
    canon
    0.64
    itures
    0.64
     Rough
    0.62
     predominantly
    0.61
    swing
    0.61
     majorities
    0.61
    Activations Density 0.029%

    No Known Activations