© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 12222
    Prev
    Next
    INDEX
    Explanations

    route definitions

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

    This neuron spots code‐style tokens—especially Angular Forms API identifiers and method-call syntax (e.g. class names, validator names, and dotted method invocations).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    sruhe
    0.55
    Geolocation
    0.54
     powierzchni
    0.54
    bildung
    0.49
     Cay
    0.49
    THA
    0.49
    מצא
    0.48
    堀
    0.48
     yng
    0.48
    انه
    0.47
    POSITIVE LOGITS
    load
    0.64
     guard
    0.63
     pathMatch
    0.63
     Guard
    0.61
    guard
    0.59
     GUARD
    0.59
    GUARD
    0.56
     guards
    0.56
     Games
    0.55
     guarding
    0.55
    Activations Density 0.001%

    No Known Activations