© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 8989
    Prev
    Next
    INDEX
    Explanations

    daily, Torch, kill, common, layer

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

    This neuron activates on numeric tokens, especially multi-digit and decimal numbers (e.g., volume, page, and statistic values).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     upholstered
    0.92
    鐮
    0.87
     danam
    0.82
     wren
    0.81
     feasts
    0.78
     psychiatrists
    0.78
     conscient
    0.77
    য়া
    0.77
     colitis
    0.77
     представителей
    0.76
    POSITIVE LOGITS
    鍰
    0.78
     Spezial
    0.75
    )#
    0.73
     Fer
    0.70
     Sı
    0.69
     В
    0.68
    town
    0.68
    がる
    0.68
     серед
    0.68
    зал
    0.67
    Activations Density 0.001%

    No Known Activations