© 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. 4866
    Prev
    Next
    INDEX
    Explanations

    pronouns referring to a group or individuals

    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
    +0.26
    1.2%
    481
    +0.15
    0.7%
    480
    +0.14
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    492
    +0.26
    0.03
    464
    +0.15
    0.03
    256
    +0.14
    0.03
    Negative Logits
     Azerb
    -0.70
     Allied
    -0.65
     Woodward
    -0.65
     Argon
    -0.64
     Berk
    -0.64
     Dres
    -0.64
     Berkshire
    -0.63
     Grail
    -0.63
     Borders
    -0.63
     Kee
    -0.62
    POSITIVE LOGITS
    quickShipAvailable
    0.90
     ][
    0.84
    amins
    0.81
    pec
    0.79
    mosp
    0.79
    Í
    0.79
    ools
    0.78
    redients
    0.78
    onut
    0.73
     ));
    0.71
    Activations Density 0.072%

    No Known Activations