© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 75345
    Prev
    Next
    INDEX
    Explanations

    prayers go

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

    This neuron fires on clause-linking tokens—especially commas and coordinating or introductory words like “and” or “In” that mark the start of a new clause.

    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
    Datas
    -1.11
     andre
    -1.05
    
    -1.02
    stackoverflow
    -1.01
     then
    -0.98
     its
    -0.98
    ⎛
    -0.96
    周泽楷
    -0.96
     języ
    -0.96
    ccn
    -0.96
    POSITIVE LOGITS
     spéciaux
    1.09
     récente
    1.07
    Another
    1.05
    лизм
    1.05
     inappropriés
    1.02
    ONESIA
    1.00
     our
    1.00
     we
    0.97
     שלנו
    0.97
    //});
    0.96
    Activations Density 0.005%

    No Known Activations