© 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. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129584
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on programming‐ or math‐style tokens—things like template/typename keywords, angle‐bracket delimiters, parentheses, and other code or formal notation elements.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    difference between

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ξη
    -0.88
    ywna
    -0.86
     cultivators
    -0.85
    Ѓ
    -0.84
    ultier
    -0.80
     vultures
    -0.80
    There
    -0.78
    Hash
    -0.77
    stopp
    -0.77
    AutoSize
    -0.76
    POSITIVE LOGITS
     other
    1.04
    other
    0.78
     filha
    0.78
     سایر
    0.76
    其他
    0.75
     ward
    0.75
    勾
    0.74
     simply
    0.73
     further
    0.72
    confirmButton
    0.72
    Activations Density 0.099%

    No Known Activations