© 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. 98621
    Prev
    Next
    INDEX
    Explanations

    numbers and foreign entities

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

    The neuron strongly activates on numeric tokens and their surrounding whitespace (i.e. digit sequences in numbers).

    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
     as
    -1.23
     now
    -1.18
     does
    -1.17
     adding
    -1.16
     its
    -1.15
     going
    -1.15
     both
    -1.14
     go
    -1.14
     what
    -1.13
     (
    -1.12
    POSITIVE LOGITS
     welkom
    1.17
    beispiel
    1.16
    thisis
    1.16
     délib
    1.13
     auront
    1.10
    柔ら
    1.06
     incrí
    1.05
     innumer
    1.05
     konkuren
    1.04
    柔らかい
    1.03
    Activations Density 0.021%

    No Known Activations