© 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. 21011
    Prev
    Next
    INDEX
    Explanations

    phrases related to marine life and environments

    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
    +0.13
    0.6%
    52
    +0.12
    0.5%
    548
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.13
    0.01
    288
    +0.12
    0.02
    548
    +0.10
    0.01
    Negative Logits
    Interstitial
    -0.83
     CBI
    -0.75
     Eliot
    -0.75
    é¾įåĸļ士
    -0.74
    Madison
    -0.73
    ourney
    -0.71
    ãĥ´
    -0.70
    Charges
    -0.70
    UGC
    -0.69
    GV
    -0.69
    POSITIVE LOGITS
     reef
    1.28
     reefs
    1.18
     aquarium
    1.08
     habitat
    0.97
     Reef
    0.97
    fish
    0.96
     algae
    0.94
     fisher
    0.91
     coral
    0.87
     larvae
    0.86
    Activations Density 0.025%

    No Known Activations