© 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. 3061
    Prev
    Next
    INDEX
    Explanations

    names and titles of people

    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
    659
    +0.24
    1.1%
    547
    +0.09
    0.4%
    508
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.24
    0.13
    103
    +0.09
    0.06
    735
    +0.09
    0.05
    Negative Logits
     MSG
    -0.81
    ModLoader
    -0.74
     TVs
    -0.71
     neon
    -0.67
    knit
    -0.66
    Women
    -0.66
     GST
    -0.65
     fridge
    -0.64
     governing
    -0.63
    FACE
    -0.63
    POSITIVE LOGITS
    linger
    1.05
    isson
    1.05
    essler
    1.01
    isner
    0.94
    agan
    0.94
    ould
    0.93
    antz
    0.93
    .,
    0.92
    iggins
    0.92
    ohn
    0.92
    Activations Density 0.113%

    No Known Activations