© 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. 3776
    Prev
    Next
    INDEX
    Explanations

    any words with the "ex" prefix

    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
    463
    +0.11
    0.5%
    208
    +0.10
    0.5%
    115
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    314
    +0.11
    0.01
    208
    +0.10
    0.03
    463
    +0.10
    0.03
    Negative Logits
    ©¶æ¥µ
    -0.92
    STER
    -0.77
    WOOD
    -0.70
     Frey
    -0.67
    HAEL
    -0.65
    è£ħ
    -0.64
    SHARE
    -0.64
    Ĥİ
    -0.63
    FUL
    -0.63
    LESS
    -0.63
    POSITIVE LOGITS
    acters
    1.00
    otation
    0.93
    tenance
    0.93
    ruction
    0.89
    vised
    0.89
    mission
    0.87
    amples
    0.86
    ction
    0.85
    iliation
    0.84
    lected
    0.84
    Activations Density 0.100%

    No Known Activations