© 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 EvalsExportsAPI 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. 6287
    Prev
    Next
    INDEX
    Explanations

    mentions of the Navy, Navy SEALs, and military-related terms

    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
    288
    +0.17
    0.8%
    152
    +0.13
    0.6%
    269
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    152
    +0.17
    0.02
    269
    +0.13
    0.01
    288
    +0.12
    0.01
    Negative Logits
    DonaldTrump
    -0.72
    tered
    -0.67
    olkien
    -0.67
     Gree
    -0.67
     Gandhi
    -0.65
    otten
    -0.65
    aphael
    -0.65
     DRM
    -0.64
    payer
    -0.64
    oned
    -0.63
    POSITIVE LOGITS
     SEAL
    1.41
     Yard
    1.16
     Seal
    1.03
     Fisheries
    0.87
     Reserve
    0.86
     Corps
    0.84
     sailor
    0.82
     Sea
    0.81
    boats
    0.81
    ategory
    0.79
    Activations Density 0.010%

    No Known Activations