© 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. 20592
    Prev
    Next
    INDEX
    Explanations

    phrases related to disagreements, risks, and obstacles

    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
    447
    +0.14
    0.7%
    575
    +0.11
    0.5%
    679
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    724
    +0.14
    0.07
    575
    +0.11
    0.06
    324
    +0.10
    0.05
    Negative Logits
    bilt
    -0.68
    atorium
    -0.66
    EVA
    -0.61
    frey
    -0.61
    Benz
    -0.61
    LESS
    -0.61
    athon
    -0.60
    asus
    -0.59
    deen
    -0.59
    OV
    -0.59
    POSITIVE LOGITS
    poons
    1.09
    omething
    1.08
    pring
    1.04
     gal
    1.03
    cale
    1.01
    ensical
    1.00
    mith
    0.99
    paces
    0.97
    heet
    0.95
    pace
    0.94
    Activations Density 0.263%

    No Known Activations