© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 101192
    Prev
    Next
    INDEX
    Explanations

    verbs related to performing tasks

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

    The neuron detects imperative‐style action words used as section headers or commands (e.g. “Improve,” “Provide,” “Determine”).

    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
    鹄
    -1.98
    ＂
    -1.97
     \
    -1.92
    ൎ
    -1.90
     is
    -1.89
    好吗
    -1.89
     at
    -1.88
    紧接着
    -1.87
    잦
    -1.82
    怍
    -1.80
    POSITIVE LOGITS
    to
    3.03
    ſelf
    2.88
     uſed
    2.38
    on
    2.36
    ússia
    2.31
    ꦧ
    2.25
     глазах
    2.22
    逦
    2.20
    UREAU
    2.17
    呴
    2.13
    Activations Density 0.049%

    No Known Activations