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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 15858
    Prev
    Next
    INDEX
    Explanations

    abbreviated personal or company names

    oai_token-act-pair · gpt-3.5-turboTriggered 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
    480
    +0.54
    3.1%
    266
    +0.15
    0.9%
    87
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.54
    0.10
    232
    +0.15
    0.09
    480
    +0.15
    0.02
    Negative Logits
     worthwhile
    -0.55
     perpet
    -0.54
     credential
    -0.52
     NETWORK
    -0.51
     conducive
    -0.51
     convenient
    -0.51
     discouraging
    -0.51
     compounded
    -0.51
     backlog
    -0.51
     rewarding
    -0.51
    POSITIVE LOGITS
    ena
    0.70
    oya
    0.68
    ona
    0.67
    enne
    0.66
    itan
    0.65
    ife
    0.64
    tek
    0.64
    areth
    0.63
    ola
    0.62
    anny
    0.62
    Activations Density 0.488%

    No Known Activations