© 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. 14848
    Prev
    Next
    INDEX
    Explanations

    information about the trading, characteristics, and details of various items or concepts in the context of gaming and online activities

    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
    640
    +0.32
    1.5%
    447
    +0.18
    0.9%
    143
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.32
    0.08
    143
    +0.18
    0.04
    179
    +0.10
    0.04
    Negative Logits
    hement
    -1.01
    ilipp
    -0.70
     princ
    -0.69
    tsky
    -0.64
     precaution
    -0.64
     corresponding
    -0.63
     compan
    -0.62
     agre
    -0.61
     reflected
    -0.61
     citiz
    -0.60
    POSITIVE LOGITS
    ³³³³³³³³³³³³³³³³
    0.92
    Legendary
    0.86
    ³³³³³³³³
    0.85
    Premium
    0.84
    minster
    0.76
    Grade
    0.76
    ³³³
    0.75
    ³³³³
    0.74
    Liter
    0.72
    Wide
    0.68
    Activations Density 0.131%

    No Known Activations