© 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. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 8277
    Prev
    Next
    INDEX
    Explanations

    Si

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron selectively activates on numeric tokens—especially multi‐digit numbers and decimal measurements.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     både
    -2.20
     også
    -2.19
     bzw
    -2.16
    先是
    -2.14
    RICIA
    -2.09
    on
    -2.06
     jedoch
    -1.97
     but
    -1.97
    ↵
    -1.92
    平静
    -1.87
    POSITIVE LOGITS
     whatnot
    2.89
    蔼
    2.86
     sebagainya
    2.58
    行
    2.36
    We
    2.14
     یا
    2.14
     then
    2.11
    rogens
    2.06
    amp
    2.05
     consequent
    2.03
    Activations Density 0.057%

    No Known Activations