© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 75349
    Prev
    Next
    INDEX
    Explanations

    language followed by language name

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

    The neuron activates on metadata field labels (e.g. “Language,” “Country,” “Year,” “Gender”) in structured info sections.

    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
    !!!!
    -1.15
    </em>
    -1.13
    !!!!!
    -1.12
    !!!
    -1.11
     businesses
    -1.10
     at
    -1.09
     that
    -1.06
     
    -1.06
    ….
    -1.05
     as
    -1.02
    POSITIVE LOGITS
    Santé
    1.35
     inond
    1.28
     superbe
    1.23
    IKAN
    1.21
     VARIOUS
    1.18
     discut
    1.17
     netts
    1.17
    primarily
    1.16
     enthou
    1.16
     velg
    1.13
    Activations Density 0.021%

    No Known Activations