© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 71645
    Prev
    Next
    INDEX
    Explanations

    names, dates, or controlling actions

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

    The neuron consistently activates on content words (substantive nouns or verbs) that carry “semantic weight” rather than on function words or punctuation.

    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
    Fakta
    -1.38
     supaya
    -1.38
     sebelum
    -1.34
    Westfalen
    -1.33
    Each
    -1.30
     menyenangkan
    -1.28
    Before
    -1.27
    Significado
    -1.26
    Those
    -1.23
    Leírás
    -1.23
    POSITIVE LOGITS
    née
    1.09
    这类
    1.03
     Tienen
    0.96
     Inventories
    0.95
    poptosis
    0.95
     [...]
    0.95
    ."
    0.95
    缔
    0.93
    breviations
    0.92
    采用了
    0.92
    Activations Density 0.047%

    No Known Activations