© 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. 65879
    Prev
    Next
    INDEX
    Explanations

    foreign language phrases

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

    The neuron fires strongly on short function‐word tokens—especially prepositions or preposition+article contractions (like “at,” French “le,” Portuguese “nela”/“dela,” Hebrew clitics)—across multiple languages.

    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
    那边
    -2.03
    吶
    -1.70
     davran
    -1.69
     valla
    -1.68
    涐
    -1.66
    
    -1.65
     the
    -1.64
     ते
    -1.63
    朥
    -1.63
     ausreich
    -1.62
    POSITIVE LOGITS
     It
    2.14
     unſ
    1.98
    s
    1.87
     Someone
    1.82
     Anyone
    1.81
     що
    1.81
    大多
    1.80
     Notably
    1.78
     ſever
    1.78
    ר
    1.77
    Activations Density 0.001%

    No Known Activations