© 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. 6942
    Prev
    Next
    INDEX
    Explanations

    words related to leaks or spills

    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
    260
    +0.12
    0.5%
    65
    +0.12
    0.5%
    354
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.12
    0.01
    366
    +0.12
    0.01
    738
    +0.09
    0.01
    Negative Logits
    ĪĴ
    -0.83
     Pwr
    -0.75
    ¥ŀ
    -0.71
     tert
    -0.71
     variance
    -0.70
     trib
    -0.69
    ICAN
    -0.69
     lieu
    -0.69
     Metatron
    -0.67
     Samar
    -0.66
    POSITIVE LOGITS
    glers
    0.96
    ety
    0.91
    oslav
    0.87
    clips
    0.86
    bery
    0.81
    erville
    0.79
    wear
    0.78
    azines
    0.76
    odon
    0.75
    nos
    0.72
    Activations Density 0.031%

    No Known Activations