© 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. 33713
    Prev
    Next
    INDEX
    Explanations

    code definitions and programming terms

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

    This neuron fires on quoted string literals—particularly the human-readable names (e.g. verbose_name, related_name, parameter names) inside single quotes.

    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
     kebaya
    -1.16
     Incluye
    -1.14
    advised
    -1.10
    前は
    -1.09
    方針
    -1.08
    などを
    -1.08
     Obtén
    -1.06
    <bos>
    -1.05
    などで
    -1.05
     mutiara
    -1.05
    POSITIVE LOGITS
     Министер
    1.14
    .
    1.04
    Ꭺ
    1.02
    ItemStack
    1.01
    “
    0.91
    ādi
    0.89
     gerek
    0.89
     étrangers
    0.88
     ingenieros
    0.88
    zm
    0.88
    Activations Density 0.000%

    No Known Activations