© 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. 17885
    Prev
    Next
    INDEX
    Explanations

    words related to geographical locations, particularly ones that involve territorial disputes

    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
    420
    +0.13
    0.6%
    710
    +0.12
    0.5%
    16
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    299
    +0.13
    0.01
    582
    +0.12
    0.01
    659
    +0.11
    0.01
    Negative Logits
    ulously
    -0.74
    lished
    -0.70
    ãĤ¢ãĥ«
    -0.69
    wcsstore
    -0.66
    glomer
    -0.65
    arb
    -0.64
    ultan
    -0.64
    ulin
    -0.63
    izoph
    -0.63
     Jinping
    -0.61
    POSITIVE LOGITS
    cule
    1.35
    tto
    1.34
    cules
    1.34
    tta
    1.30
    tti
    1.24
    lette
    1.16
    cular
    1.15
    eper
    0.96
    ttes
    0.92
    mn
    0.89
    Activations Density 0.034%

    No Known Activations