© 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. 22-GEMMASCOPE-RES-131K
    4. 10819
    Prev
    Next
    INDEX
    Explanations

    parentheses and code constructs

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

    The neuron strongly fires on tokens involved in math‐style function calls or definitions—i.e. a name immediately followed by parentheses (and its argument) in formulas or code.

    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
    mentet
    -0.93
     GPR
    -0.92
    Föld
    -0.90
     उपकरण
    -0.86
     Sándor
    -0.82
    続きます
    -0.81
    водом
    -0.81
    Prés
    -0.80
    صادر
    -0.79
     khoa
    -0.79
    POSITIVE LOGITS
    ridescent
    1.00
    rits
    0.88
     добав
    0.87
     kuj
    0.84
    carus
    0.84
    puis
    0.84
    ctv
    0.82
     Voici
    0.82
    
    0.81
    THEN
    0.81
    Activations Density 0.050%

    No Known Activations