© 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. 39318
    Prev
    Next
    INDEX
    Explanations

    roles and titles

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

    The neuron activates on standalone numeric tokens—especially numbers appearing after dashes or “as” (e.g. role or rating values) in lists.

    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
     problemet
    -2.05
     mâle
    -1.95
     régal
    -1.91
     marqués
    -1.87
     gouttes
    -1.80
     périph
    -1.78
     parfüm
    -1.78
     normalt
    -1.77
     méda
    -1.77
     totalt
    -1.70
    POSITIVE LOGITS
     to
    3.06
     and
    2.50
     in
    2.31
     by
    2.11
     same
    1.86
     within
    1.81
     under
    1.81
     between
    1.79
     now
    1.77
     either
    1.77
    Activations Density 0.020%

    No Known Activations