© 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. 38827
    Prev
    Next
    INDEX
    Explanations

    asking math questions

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

    The neuron fires on words that mark the progression of solution steps (e.g. “Finding,” “gives,” “steps,” “last,” “confirmed”).

    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
    unno
    -1.75
     parámetros
    -1.74
     tabela
    -1.72
    TIPOS
    -1.72
     controladores
    -1.72
    ᾽
    -1.71
     pacote
    -1.70
     sitios
    -1.67
     información
    -1.67
     rí
    -1.66
    POSITIVE LOGITS
     with
    2.58
     also
    2.05
    ῖν
    1.99
    Also
    1.98
     one
    1.95
     дорог
    1.66
     Also
    1.66
    ῖς
    1.63
     because
    1.62
     although
    1.61
    Activations Density 1.033%

    No Known Activations