© 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. 66045
    Prev
    Next
    INDEX
    Explanations

    structural prepositions and conjunctions

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

    The neuron activates on question marks, flagging when a sentence or line is posed as a question.

    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
     rapidement
    -1.21
     pequeñas
    -1.02
    の良い
    -1.00
    のいい
    -0.99
     véritable
    -0.99
     étaient
    -0.99
     curieux
    -0.99
    Ď
    -0.98
     prestaciones
    -0.97
    镬
    -0.97
    POSITIVE LOGITS
     when
    1.41
     after
    1.17
     because
    1.16
     Because
    1.08
    when
    1.08
     if
    1.02
     to
    1.02
     using
    1.00
    했
    0.97
     on
    0.96
    Activations Density 0.087%

    No Known Activations