© 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. 10-GEMMASCOPE-RES-131K
    7. 104510
    Prev
    Next
    INDEX
    Explanations

    action verbs, emotional expression

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

    The neuron is a detector for conjugated action words—that is, inflected verbs (e.g. past‐tense “rubbed,” “grasped,” “lamented,” or third‐person present “pushes”).

    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
     a
    -1.80
     instituições
    -1.66
     Melhor
    -1.66
     mudanças
    -1.61
     which
    -1.59
     Consejos
    -1.59
     vuestros
    -1.58
     verheir
    -1.56
    々な
    -1.53
    逦
    -1.51
    POSITIVE LOGITS
    be
    1.48
     did
    1.48
     свою
    1.44
    仅
    1.41
    是
    1.36
     gets
    1.32
     enhances
    1.31
    攢
    1.30
     first
    1.29
     still
    1.26
    Activations Density 0.003%

    No Known Activations