© 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. 22-GEMMASCOPE-RES-131K
    7. 73123
    Prev
    Next
    INDEX
    Explanations

    class constructors and definitions

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

    This neuron strongly activates on C++ class (or struct) names in constructor definitions (i.e. the identifier matching its enclosing type at the start of a function definition).

    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
    backen
    -1.20
    fabri
    -1.20
     niemi
    -1.16
    organe
    -1.15
     niem
    -1.13
    のお知らせ
    -1.13
    eabouts
    -1.13
    breit
    -1.08
     baden
    -1.08
     milhões
    -1.08
    POSITIVE LOGITS
    售价
    1.01
    habited
    0.97
    Другие
    0.96
     אלה
    0.93
    ratie
    0.93
    Still
    0.92
    omyelitis
    0.91
    冊
    0.91
     Legislat
    0.91
    她们
    0.90
    Activations Density 0.005%

    No Known Activations