© 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. 34-GEMMASCOPE-RES-131K
    7. 78423
    Prev
    Next
    INDEX
    Explanations

    variable declaration, url or paths, package, or content fields

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

    This neuron fires on Chinese technical documentation terms—especially nouns related to programming, data formats, and versioning.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     weinig
    -1.19
     desselben
    -1.18
     précédemment
    -1.16
    ときは
    -1.13
    為に
    -1.10
    vermectin
    -1.09
     великолеп
    -1.05
    ようになる
    -1.05
     diesmal
    -1.05
    取决于
    -1.04
    POSITIVE LOGITS
     also
    1.86
    は
    1.77
     not
    1.77
     already
    1.57
     still
    1.57
     definitely
    1.50
     always
    1.44
     supposed
    1.41
     certainly
    1.35
     is
    1.34
    Activations Density 0.007%

    No Known Activations