© 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. 10-GEMMASCOPE-RES-131K
    7. 19852
    Prev
    Next
    INDEX
    Explanations

    as X says

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

    The neuron fires on numeric tokens—years, page or figure references, counts, and other standalone numbers in the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     also
    -1.99
     not
    -1.63
     because
    -1.59
    EDAD
    -1.46
    에도
    -1.45
    
    -1.42
    着一个
    -1.41
    =
    -1.38
     також
    -1.34
     Because
    -1.34
    POSITIVE LOGITS
    не
    1.46
     例
    1.42
    woorden
    1.38
    corrhi
    1.34
    ksesta
    1.34
     particularmente
    1.25
     morgens
    1.25
    например
    1.25
    AudioSource
    1.23
    useNewUrlParser
    1.23
    Activations Density 0.011%

    No Known Activations