© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 59239
    Prev
    Next
    INDEX
    Explanations

    words preceding other words

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

    The neuron selectively activates on common auxiliary and modal words (e.g. can, will, seems, were, going) that signal modality or tense.

    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
     for
    -2.11
     whose
    -2.11
     not
    -2.05
     who
    -1.88
     where
    -1.86
     in
    -1.84
     which
    -1.80
     if
    -1.72
     with
    -1.66
     to
    -1.64
    POSITIVE LOGITS
     داشتم
    1.59
    Tableau
    1.56
    ٹ
    1.53
     huiles
    1.46
    🔺
    1.45
    pendente
    1.43
    انگلیسی
    1.42
     mélang
    1.41
     péné
    1.41
    他們
    1.41
    Activations Density 0.086%

    No Known Activations