© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 3864
    Prev
    Next
    INDEX
    Explanations

    phrases beginning with the word "the"

    oai_token-act-pair · gpt-3.5-turbo
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    362
    +0.22
    1.1%
    393
    +0.17
    0.8%
    16
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.22
    0.14
    16
    +0.17
    0.09
    327
    +0.13
    0.08
    Negative Logits
    arians
    -0.70
    ï¸ı
    -0.69
    ibl
    -0.69
    load
    -0.69
    pointers
    -0.68
    handedly
    -0.67
    strap
    -0.63
    ional
    -0.62
    ulence
    -0.62
     declares
    -0.61
    POSITIVE LOGITS
     midst
    1.36
     guise
    1.22
     meantime
    1.21
     vicinity
    1.20
     aftermath
    1.18
     Philippines
    1.12
     same
    1.09
     Netherlands
    1.03
     wake
    1.02
     Bahamas
    1.01
    Activations Density 0.301%

    No Known Activations