© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 13084
    Prev
    Next
    INDEX
    Explanations

    years, dates

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

    This neuron is primarily activated by four-digit numeric tokens, especially years (e.g., “2016,” “2017,” etc.).

    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_31_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
     ২০০
    0.75
    쨋
    0.61
     किन्तु
    0.57
    ศัพท์
    0.57
    0
    0.57
     ২২
    0.56
    ృతి
    0.54
    ０
    0.54
     ২৩
    0.54
    ২০০
    0.53
    POSITIVE LOGITS
     manier
    0.66
     so
    0.64
     facere
    0.62
     soooo
    0.61
     bastante
    0.60
     sooo
    0.58
     так
    0.55
     partes
    0.54
     tellement
    0.54
     regio
    0.54
    Activations Density 0.142%

    No Known Activations