© 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. 4634
    Prev
    Next
    INDEX
    Explanations

    adjectives like "award-winning" or related phrases

    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
    120
    +0.12
    0.5%
    314
    +0.11
    0.5%
    407
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    152
    +0.12
    0.04
    120
    +0.11
    0.04
    124
    +0.10
    0.03
    Negative Logits
     Ukrain
    -0.80
     spont
    -0.73
     Silk
    -0.67
     Peoples
    -0.67
     electroly
    -0.67
     skate
    -0.66
     proxies
    -0.66
     Cree
    -0.65
     OC
    -0.65
    igion
    -0.65
    POSITIVE LOGITS
    winning
    1.31
    eligible
    1.23
    favorite
    1.08
    laden
    1.07
    sized
    1.00
    derived
    0.97
    protected
    0.97
    caliber
    0.96
    bearing
    0.95
    rated
    0.95
    Activations Density 0.072%

    No Known Activations