© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 25045
    Prev
    Next
    INDEX
    Explanations

    is/are/was/were + participle

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

    The neuron strongly activates on the core predicate of a sentence (the main verb or copula plus its surrounding high-frequency function words like “to,” “a,” and “is”).

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     tutur
    -1.27
     publicités
    -1.19
    龇
    -1.17
    jelaskan
    -1.12
    ungkapkan
    -1.10
     calendriers
    -1.09
     държа
    -1.08
     poissons
    -1.08
    specifically
    -1.07
    ebly
    -1.06
    POSITIVE LOGITS
     will
    1.32
     have
    1.32
    之一
    1.15
    ábor
    1.14
    olesterol
    1.10
    
    1.09
     has
    1.09
     helping
    1.07
     didn
    1.05
    ">
    1.05
    Activations Density 0.054%

    No Known Activations