© 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. 14839
    Prev
    Next
    INDEX
    Explanations

    numerical values related to game statistics or attributes

    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
    71
    +0.12
    0.5%
    709
    +0.10
    0.5%
    119
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.12
    0.01
    255
    +0.10
    0.02
    375
    +0.10
    0.01
    Negative Logits
    sonian
    -0.89
    urn
    -0.77
     Bland
    -0.76
    ovie
    -0.73
    arte
    -0.71
    rial
    -0.69
    uthor
    -0.69
    iment
    -0.68
     Borders
    -0.68
    orld
    -0.68
    POSITIVE LOGITS
     inflicted
    0.93
     modifier
    0.90
     dealt
    0.90
     multiplier
    0.89
    damage
    0.89
     damage
    0.87
     spikes
    0.86
     modifiers
    0.86
     reduction
    0.84
     mitigation
    0.83
    Activations Density 0.038%

    No Known Activations