© 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. 14859
    Prev
    Next
    INDEX
    Explanations

    math equations and code

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

    The neuron fires on mathematical notation—especially numbers, formulas, and function names (e.g. Sin, Cos, Table).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
    ּוֹ
    -1.34
    וּ
    -1.18
    Dimensões
    -1.12
     verhind
    -1.07
     nimmer
    -1.06
    }}_{\
    -1.03
    quartier
    -1.03
     jengibre
    -1.02
    gehe
    -1.01
    ınız
    -1.00
    POSITIVE LOGITS
     ویتامین
    1.09
     where
    1.08
    ׇ
    1.04
     چربی
    0.97
     also
    0.96
     گیاه
    0.96
     ironically
    0.94
    ὴν
    0.94
     some
    0.93
     avoiding
    0.93
    Activations Density 0.009%

    No Known Activations