© 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. 66466
    Prev
    Next
    INDEX
    Explanations

    mathematical questions or statements

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

    The neuron overwhelmingly lights up on numeric literals (tokens composed of digits, e.g. “0”, “120”, “202”) rather than on letters or operators.

    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
     but
    -0.98
     основ
    -0.93
    Prem
    -0.88
    琉璃
    -0.85
     GOLDEN
    -0.83
     JOSÉ
    -0.83
     minut
    -0.81
    ⮚
    -0.80
    ##
    -0.80
     barbara
    -0.79
    POSITIVE LOGITS
    values
    1.05
     ราคา
    0.99
    ixin
    0.98
     ludzie
    0.96
     yapıl
    0.95
    讥
    0.95
    Typeface
    0.93
     tambin
    0.93
    helves
    0.93
     Isso
    0.91
    Activations Density 0.001%

    No Known Activations