© 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. 10-GEMMASCOPE-RES-131K
    7. 84081
    Prev
    Next
    INDEX
    Explanations

    commands and transitions

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

    The neuron strongly activates on base‐form verbs used as commands or directives (i.e. imperatives).

    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
     or
    -1.48
    Namun
    -1.37
     also
    -1.36
    さんも
    -1.35
     like
    -1.32
     helping
    -1.30
    を目指す
    -1.29
    webtoken
    -1.28
    Mereka
    -1.27
    elancer
    -1.27
    POSITIVE LOGITS
    いいのか
    1.39
     simplement
    1.38
     汽车
    1.34
     刘
    1.30
     できる
    1.29
    踯
    1.27
     verschiedener
    1.26
    icona
    1.25
     his
    1.24
     “
    1.22
    Activations Density 0.051%

    No Known Activations