© 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. 5160
    Prev
    Next
    INDEX
    Explanations

    mentions of news agencies and official statements

    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
    725
    +0.14
    0.6%
    506
    +0.10
    0.5%
    547
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    115
    +0.14
    0.04
    725
    +0.10
    0.04
    267
    +0.10
    0.04
    Negative Logits
    otin
    -0.91
     Lovecraft
    -0.82
    weed
    -0.81
    etsy
    -0.81
    gregation
    -0.80
    abiding
    -0.69
     Woods
    -0.68
     Deck
    -0.68
     Millennials
    -0.67
    ophers
    -0.67
    POSITIVE LOGITS
    ews
    1.10
     quoting
    0.92
     news
    0.88
     quoted
    0.87
     correspondent
    0.87
     newspaper
    0.86
     outlets
    0.84
     sources
    0.83
     translated
    0.83
     quotes
    0.83
    Activations Density 0.156%

    No Known Activations