© 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. 12825
    Prev
    Next
    INDEX
    Explanations

    keywords related to Korean affairs and politics

    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
    106
    +0.13
    0.6%
    302
    +0.11
    0.5%
    558
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    302
    +0.13
    0.02
    67
    +0.11
    0.02
    106
    +0.10
    0.02
    Negative Logits
    ashtra
    -0.76
     Mayweather
    -0.76
    llo
    -0.76
    MENTS
    -0.76
    instein
    -0.75
    clair
    -0.75
    yard
    -0.74
    gerald
    -0.70
    ttes
    -0.70
    MENT
    -0.70
    POSITIVE LOGITS
    orea
    1.15
     peninsula
    1.08
     ì
    1.02
     Korea
    0.97
     ë
    0.97
     Peninsula
    0.93
    ë
    0.90
     Koreans
    0.84
     Jong
    0.83
     Korean
    0.83
    Activations Density 0.023%

    No Known Activations