© 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. 22053
    Prev
    Next
    INDEX
    Explanations

    references to issues or editions of publications

    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
    393
    +0.12
    0.5%
    373
    +0.10
    0.5%
    572
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    572
    +0.12
    0.03
    362
    +0.10
    0.03
    587
    +0.10
    0.03
    Negative Logits
    cffffcc
    -0.77
    headed
    -0.74
    dinand
    -0.69
     nod
    -0.69
    anders
    -0.69
    istas
    -0.69
     wiser
    -0.67
     edges
    -0.67
    etooth
    -0.67
     clasp
    -0.67
    POSITIVE LOGITS
     Premiere
    0.85
     Weekly
    0.84
     ABC
    0.81
     Monthly
    0.80
     Forbes
    0.80
     Trend
    0.75
     Notting
    0.75
     PBS
    0.75
     Lantern
    0.74
     Variety
    0.74
    Activations Density 0.121%

    No Known Activations