© 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. 17017
    Prev
    Next
    INDEX
    Explanations

    phrases related to controversial or sensitive topics, such as race, gender, and social issues

    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
    330
    +0.10
    0.5%
    620
    +0.09
    0.4%
    225
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    251
    +0.10
    0.04
    232
    +0.09
    0.05
    326
    +0.09
    0.04
    Negative Logits
     Clause
    -0.72
    FUN
    -0.65
     Georgian
    -0.58
     chronological
    -0.57
    sonian
    -0.57
     Dragonbound
    -0.57
     Window
    -0.57
     Evangel
    -0.56
     Abyss
    -0.55
     Geo
    -0.55
    POSITIVE LOGITS
    rolled
    1.02
    imposed
    1.01
    ersed
    0.97
    ared
    0.92
    inged
    0.90
    aped
    0.90
    overed
    0.88
    ighed
    0.87
    assed
    0.86
    turned
    0.85
    Activations Density 0.258%

    No Known Activations