© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 76390
    Prev
    Next
    INDEX
    Explanations

    quantifiers before things

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

    The neuron activates on words that quantify or enumerate (e.g. numbers and quantifier adjectives like “two,” “only,” “many,” “different”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    ']->
    -1.08
    );
    
    -0.95
    zenio
    -0.95
     какое
    -0.94
    ámica
    -0.94
    )));
    
    -0.94
     Borja
    -0.93
    ]);
    
    -0.93
     stadion
    -0.92
    )=\
    -0.91
    POSITIVE LOGITS
     things
    9.25
     thing
    5.59
    Things
    5.50
    things
    5.13
     Things
    5.06
     cosas
    5.06
     coisas
    4.66
     THINGS
    4.22
    THINGS
    3.94
     dingen
    3.89
    Activations Density 0.167%

    No Known Activations