© 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. 20339
    Prev
    Next
    INDEX
    Explanations

    more than half

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

    The neuron strongly activates on numeric quantity expressions—especially proportions or percentages (e.g. “half,” “two-thirds,” “%” and similar fraction/percentage tokens).

    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
     maybe
    -1.67
     perhaps
    -1.65
     or
    -1.64
    谕
    -1.41
    0
    -1.40
    ↵↵
    -1.38
     Hiring
    -1.34
    しつつ
    -1.34
    过来了
    -1.33
     about
    -1.32
    POSITIVE LOGITS
    StandardCharsets
    1.63
     kommentar
    1.59
    Cruise
    1.52
    臵
    1.49
     ledig
    1.47
    mismo
    1.47
    autorelease
    1.45
     grupp
    1.45
    genodigd
    1.44
    mdan
    1.44
    Activations Density 0.023%

    No Known Activations