© 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. 31-GEMMASCOPE-2-RES-16K
    7. 1540
    Prev
    Next
    INDEX
    Explanations

    multi and dual concepts

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

    This neuron detects multiplicative prefixes or indicators of quantity (e.g. “double,” “dual,” “triple,” “multi-”) in words.

    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
     Astrology
    0.57
     nato
    0.54
    σ
    0.54
     sofr
    0.53
    IRST
    0.53
    sj
    0.53
    ONLY
    0.52
     zu
    0.51
     اجمع
    0.51
    brought
    0.51
    POSITIVE LOGITS
    pronged
    1.06
     sided
    1.04
     layered
    1.00
    双
    0.95
     camadas
    0.93
     prong
    0.90
     layers
    0.86
     triple
    0.86
    layered
    0.86
    雙
    0.86
    Activations Density 0.375%

    No Known Activations