© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 66221
    Prev
    Next
    INDEX
    Explanations

    class name declarations

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

    The neuron fires on code constructs that define or declare types—especially the “class” and related keywords (like “protocol” or “extends”).

    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
     expédié
    -1.25
     Durée
    -1.19
    gencias
    -1.17
    mazione
    -1.10
    randint
    -1.09
    ffy
    -1.09
    áculo
    -1.06
    aaS
    -1.06
    UTION
    -1.05
    kym
    -1.05
    POSITIVE LOGITS
    下車
    1.04
    などを
    0.98
    8
    0.90
     courseId
    0.89
    などの
    0.88
    '
    0.88
     такі
    0.88
     الجلد
    0.88
     pendukung
    0.88
     şeklinde
    0.87
    Activations Density 0.006%

    No Known Activations