© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 67671
    Prev
    Next
    INDEX
    Explanations

    commands like add, delete, update, install, reset

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

    The neuron fires on programming identifiers that are imperative action verbs (e.g. “read,” “free,” “instantiate,” “update,” “remove,” “disable,” etc.) in code, i.e. function or method names beginning with a verb.

    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
    unno
    -1.31
    APRIL
    -1.23
    鸲
    -1.20
    机遇
    -1.20
     informace
    -1.20
     ILLUSTRATIONS
    -1.19
    ਗ਼
    -1.19
    ridescent
    -1.16
    
    -1.13
    pherds
    -1.12
    POSITIVE LOGITS
     of
    1.91
     them
    1.44
     your
    1.29
     different
    1.27
     a
    1.09
     already
    1.09
     each
    1.08
     potuto
    1.05
     serta
    1.05
     very
    1.01
    Activations Density 0.008%

    No Known Activations