© 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. 22-GEMMASCOPE-RES-131K
    7. 25449
    Prev
    Next
    INDEX
    Explanations

    North America

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

    The neuron strongly activates on common high‐frequency function words (e.g. pronouns, conjunctions, prepositions, question words) used to connect or introduce clauses.

    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
     luf
    -1.20
    ۴
    -1.11
     situe
    -1.10
     Sosial
    -1.09
     biograf
    -1.06
    ۷
    -1.05
     menyen
    -1.05
    ۶
    -1.05
    anderen
    -1.03
     disiplin
    -1.03
    POSITIVE LOGITS
     until
    1.16
     llevará
    1.03
     different
    1.01
     verschillen
    0.98
     before
    0.97
     strongly
    0.97
     immensely
    0.95
     huge
    0.94
     make
    0.94
     describing
    0.93
    Activations Density 0.233%

    No Known Activations