© 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. 22-GEMMASCOPE-RES-131K
    7. 104634
    Prev
    Next
    INDEX
    Explanations

    introduction section headers

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

    The neuron selectively fires on numeric tokens (digits), such as those in DOIs, reference numbers, and section or figure labels.

    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
     teto
    -1.15
     manquante
    -1.11
    lacing
    -1.08
     considérée
    -1.08
     péro
    -1.07
     indiví
    -1.07
     errado
    -1.03
     établie
    -1.02
     construite
    -1.01
    髕
    -1.01
    POSITIVE LOGITS
     and
    1.16
     Beschreibung
    1.04
     to
    0.93
    了过去
    0.88
     Lainnya
    0.88
     precaution
    0.88
     jednou
    0.85
    OrderInfo
    0.85
    ség
    0.85
     gepubliceerd
    0.84
    Activations Density 0.022%

    No Known Activations