© 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. 22061
    Prev
    Next
    INDEX
    Explanations

    registration is fast

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

    The neuron fires on copular verbs (“is,” “are,” etc.) in service or policy statements announcing availability, requirements, or definitions.

    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
     of
    -2.11
     which
    -1.84
     Somehow
    -1.83
     Some
    -1.82
     with
    -1.77
     слишком
    -1.76
     Perhaps
    -1.70
     Initially
    -1.68
     what
    -1.67
     how
    -1.67
    POSITIVE LOGITS
    蟊
    1.86
    狴
    1.83
     suivre
    1.70
    
    1.65
     partager
    1.64
    
    1.64
     juſ
    1.61
     champignons
    1.60
    knowing
    1.59
    ཥ
    1.59
    Activations Density 0.110%

    No Known Activations