© 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. 22-GEMMASCOPE-RES-131K
    7. 16026
    Prev
    Next
    INDEX
    Explanations

    purposeful action verbs

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

    The neuron detects language describing formal institutional or legislative actions—e.g. approvals, votes, authorizations, recommendations, and similar decision-making events.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     seguramente
    -1.24
     quizá
    -1.18
     necesitar
    -1.17
     decisiones
    -1.16
     your
    -1.14
    麾
    -1.11
     suele
    -1.09
     might
    -1.07
     or
    -1.04
     simplemente
    -1.03
    POSITIVE LOGITS
     Médico
    1.23
     offerte
    1.11
     publie
    1.08
     kiti
    1.07
     bénéficiaire
    1.07
     secrétaire
    1.07
     époux
    1.07
     député
    1.06
    感のある
    1.03
     officier
    1.03
    Activations Density 0.056%

    No Known Activations