© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 13279
    Prev
    Next
    INDEX
    Explanations

    numbers, math, and symbols

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

    The neuron selectively activates on tokens that are numeric values (including integers, decimals, fractions, and signed numbers).

    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
    complish
    -1.17
    ditionally
    -1.11
    providedIn
    -1.04
    []);
    -0.99
    published
    -0.98
    obut
    -0.96
    }}>
    -0.95
    pearance
    -0.94
    canActivate
    -0.94
     гостиной
    -0.94
    POSITIVE LOGITS
     these
    1.02
    Ȝ
    0.92
    2
    0.91
     can
    0.84
     этих
    0.82
    newGame
    0.81
    0
    0.77
    5
    0.77
     what
    0.77
    زمة
    0.77
    Activations Density 0.068%

    No Known Activations