© 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. 101532
    Prev
    Next
    INDEX
    Explanations

    foreign languages

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

    The neuron fires strongly on common short function words—especially coordinating conjunctions (e.g. “and,” “as”) and certain prepositions (e.g. “for,” “are”).

    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
     poate
    -1.39
     be
    -1.30
    の中に
    -1.24
    を追加
    -1.22
    かっ
    -1.20
    ed
    -1.19
    -1.18
    ământ
    -1.17
    男朋友
    -1.16
     dál
    -1.14
    POSITIVE LOGITS
     from
    1.68
     because
    1.38
    dresser
    1.36
    IORS
    1.31
    五十
    1.31
     vakantie
    1.31
     Privacidad
    1.27
    suche
    1.27
     broderie
    1.27
    --

    1.27
    Activations Density 0.296%
    No Known Activations