© 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. 10-GEMMASCOPE-RES-131K
    7. 7330
    Prev
    Next
    INDEX
    Explanations

    keywords in code declarations

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

    This neuron is selectively firing on the “(” tokens that mark a constructor or method‐call argument list in Java code (e.g. the “(” after `new JButton`).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     is
    -2.09
     went
    -1.98
     came
    -1.94
     may
    -1.88
     took
    -1.84
     began
    -1.84
     started
    -1.83
     had
    -1.81
    They
    -1.80
     probably
    -1.79
    POSITIVE LOGITS
     概要
    1.66
     empêcher
    1.60
     utiliser
    1.60
     ij
    1.57
     louer
    1.56
    bereitung
    1.55
     Tô
    1.55
     arrêter
    1.54
     éve
    1.54
     asa
    1.54
    Activations Density 0.008%

    No Known Activations