© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 28024
    Prev
    Next
    INDEX
    Explanations

    lists separated by commas or conjunctions

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

    This neuron detects interface labels or command names—short words that name fields, options, or actions (e.g. menu items, search‐filter fields, image-manipulation commands).

    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
     vermel
    -1.30
     kulinar
    -1.22
     sekund
    -1.20
    ösen
    -1.20
    bakter
    -1.17
    dekor
    -1.16
     Słow
    -1.15
     portefeuille
    -1.14
     Kategor
    -1.14
    THEN
    -1.14
    POSITIVE LOGITS
     if
    1.41
     even
    1.33
     just
    1.24
     various
    1.23
     now
    1.20
     has
    1.20
     allows
    1.13
     currently
    1.13
    diteur
    1.12
     any
    1.11
    Activations Density 0.053%

    No Known Activations