© 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 49152
    9. 8-RES_FS49152-JB
    11. 41490
    Prev
    Next
    INDEX
    Explanations

    references to specific named entities or products

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    659
    +0.14
    0.7%
    481
    +0.14
    0.6%
    314
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.14
    0.10
    213
    +0.14
    0.04
    314
    +0.11
    0.01
    Negative Logits
     Marketable
    -0.70
    assetsadobe
    -0.69
    lihood
    -0.67
     wide
    -0.66
     GOODMAN
    -0.66
     EEG
    -0.63
     Tradable
    -0.63
    ģĸ
    -0.62
     Defensive
    -0.60
     widest
    -0.59
    POSITIVE LOGITS
    umble
    1.00
    ipp
    0.98
    orig
    0.92
    appy
    0.91
    argon
    0.91
    ounded
    0.89
    irc
    0.89
    aed
    0.89
    arma
    0.87
    PD
    0.86
    Activations Density 0.116%

    No Known Activations