© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 24221
    Prev
    Next
    INDEX
    Explanations

    references to websites

    oai_token-act-pair · gpt-4o-miniTriggered 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
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    373
    +0.12
    0.6%
    435
    +0.10
    0.5%
    497
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    398
    +0.12
    0.01
    435
    +0.10
    0.01
    76
    +0.09
    0.01
    Negative Logits
    rament
    -0.64
    picking
    -0.64
    DN
    -0.63
     sounding
    -0.60
    nered
    -0.59
     succeeding
    -0.59
    GBT
    -0.58
    MQ
    -0.57
     impending
    -0.57
    rous
    -0.56
    POSITIVE LOGITS
     Thumbnails
    0.73
     Flavoring
    0.69
     Gutenberg
    0.66
     Store
    0.64
    izen
    0.63
    anasia
    0.63
    CLOSE
    0.62
    ichita
    0.60
     Maintenance
    0.60
    close
    0.59
    Activations Density 0.013%

    No Known Activations