© 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 3072
    9. 8-RES_FS3072-JB
    11. 249
    Prev
    Next
    INDEX
    Explanations

    positive sentiments and expressions of gratitude

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    348
    +0.09
    0.4%
    483
    +0.09
    0.4%
    581
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    49
    +0.09
    0.09
    762
    +0.09
    0.09
    431
    +0.09
    0.07
    Negative Logits
     WWF
    -0.66
    ancel
    -0.61
     Greenpeace
    -0.59
    ammy
    -0.57
     mediation
    -0.56
     Griffin
    -0.56
     Emin
    -0.53
    winner
    -0.53
    ouses
    -0.53
    erman
    -0.52
    POSITIVE LOGITS
     Magicka
    0.69
    ":[
    0.64
    Bi
    0.63
    ."[
    0.61
    bler
    0.60
     dracon
    0.59
    enth
    0.59
     physically
    0.59
    igr
    0.58
    cells
    0.58
    Activations Density 1.220%

    No Known Activations