© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 14046
    Prev
    Next
    INDEX
    Explanations

    questions starting with "what"

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

    The neuron fires on the very first non‐trivial token in each new text segment, essentially marking the start of a snippet.

    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
     ofrec
    1.04
     நியூ
    0.99
     キ
    0.98
     vzt
    0.91
    ഗ്ര
    0.88
     값
    0.88
     き
    0.88
     décro
    0.86
     루
    0.85
    effetto
    0.85
    POSITIVE LOGITS
     Rv
    0.62
    就好
    0.61
     Disk
    0.61
    yuk
    0.61
     Phòng
    0.61
    س
    0.61
     Brig
    0.61
     Bang
    0.60
     Humber
    0.60
     Fam
    0.59
    Activations Density 0.000%

    No Known Activations