© 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. 65905
    Prev
    Next
    INDEX
    Explanations

    action followed by object

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

    The neuron detects imperative verbs that open step‐by‐step instructions.

    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
     these
    -1.27
     further
    -1.23
     it
    -1.19
     later
    -1.19
     each
    -1.18
     also
    -1.11
     such
    -1.09
     apoio
    -1.04
    naa
    -1.03
    oczes
    -1.02
    POSITIVE LOGITS
     your
    1.41
     or
    1.13
    гию
    1.09
     unwavering
    1.05
     либо
    1.05
     on
    1.05
     обычно
    1.05
     любы
    1.04
    ）　
    0.99
     entweder
    0.97
    Activations Density 0.074%

    No Known Activations