© 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. 26034
    Prev
    Next
    INDEX
    Explanations

    code descriptions and keys

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

    This neuron fires on tokens that introduce definitions or assignments—e.g. declaration keywords (static, public, module.exports, REGISTER_OP), variable/property names (widgetName, widgetIcon), front-matter keys (layout, title, image), and numeric or string literal assignments.

    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
     from
    -1.09
     ネックレス
    -0.94
    Constants
    -0.94
    čila
    -0.93
    sieve
    -0.93
    habitude
    -0.91
     жидкости
    -0.91
     biograf
    -0.91
    бливості
    -0.90
     Temmuz
    -0.89
    POSITIVE LOGITS
     description
    1.23
     instala
    1.16
     entreten
    1.16
     promete
    1.14
    要望
    1.07
     esquer
    1.06
     afec
    1.04
     creato
    1.03
     descrit
    1.03
     instructions
    1.03
    Activations Density 0.050%

    No Known Activations