© 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. 67586
    Prev
    Next
    INDEX
    Explanations

    placeholder text sequence

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

    The neuron is sharply activated by numeric tokens (it “reads” numbers in the text), outputting a nonzero response only on those digit‐only 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_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
    ového
    -1.08
    phòng
    -1.05
    ЕРЕ
    -1.04
     consulté
    -1.00
     komfort
    -0.99
     hini
    -0.99
     balón
    -0.98
    karena
    -0.98
     xadrez
    -0.98
    řízen
    -0.96
    POSITIVE LOGITS
     ebenfalls
    1.20
     but
    1.02
     But
    0.99
     again
    0.90
     cresc
    0.88
    ba
    0.88
     back
    0.87
    となる
    0.86
     encontrar
    0.86
     Again
    0.85
    Activations Density 0.001%

    No Known Activations