© 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. 22017
    Prev
    Next
    INDEX
    Explanations

    references to specific book chapters and page numbers

    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
    373
    +0.20
    0.9%
    358
    +0.11
    0.5%
    314
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    18
    +0.20
    0.02
    373
    +0.11
    -0.01
    358
    +0.11
    0.01
    Negative Logits
    ndra
    -0.66
     overnight
    -0.59
     awoke
    -0.57
     Chow
    -0.56
     Telescope
    -0.56
     Clicker
    -0.56
     Tribunal
    -0.55
     settled
    -0.55
    ministic
    -0.55
     imperson
    -0.54
    POSITIVE LOGITS
     seq
    1.00
     pg
    0.91
     pp
    0.89
     cx
    0.88
     348
    0.88
     229
    0.87
     cit
    0.86
     334
    0.85
     iv
    0.85
     324
    0.85
    Activations Density 0.049%

    No Known Activations