© 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. 95802
    Prev
    Next
    INDEX
    Explanations

    code and questions

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

    This neuron fires on Java code tokens—especially import/java keywords, type names like String, and method/accessor signatures (e.g. “get;”).

    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
     hongos
    -1.40
     sacré
    -1.22
    CURLOPT
    -1.19
     sezonu
    -1.17
     situe
    -1.17
     tendre
    -1.16
     tenues
    -1.14
     hierbas
    -1.14
     disponibilités
    -1.13
     magie
    -1.13
    POSITIVE LOGITS
     is
    1.22
     required
    1.15
     only
    1.12
     should
    1.09
     before
    1.05
    手工
    1.04
     during
    1.00
     Startseite
    0.99
     related
    0.98
     provided
    0.97
    Activations Density 0.000%

    No Known Activations