© 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. 4658
    Prev
    Next
    INDEX
    Explanations

    references related to health conditions, specific physical traits, and professional terminology

    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
    480
    +0.45
    2.2%
    532
    +0.10
    0.5%
    314
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.45
    0.08
    480
    +0.10
    0.01
    266
    +0.10
    0.06
    Negative Logits
     Lerner
    -0.70
     oneself
    -0.69
     themselves
    -0.67
    hovah
    -0.64
     guiName
    -0.63
     Kov
    -0.61
     extrad
    -0.61
     Canaver
    -0.60
     Christensen
    -0.60
     Juven
    -0.59
    POSITIVE LOGITS
     counterparts
    1.06
     brethren
    0.99
     buddies
    0.99
     counterpart
    0.97
    mates
    0.96
     mates
    0.92
     colleague
    0.88
    mate
    0.87
     instincts
    0.86
     skills
    0.83
    Activations Density 0.475%

    No Known Activations