© 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. 24-GEMMASCOPE-2-RES-16K
    7. 3814
    Prev
    Next
    INDEX
    Explanations

    that explains

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

    The neuron activates strongly on numeric expressions (numbers, decimals, measurements).

    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
     reinvigor
    0.51
    ০০
    0.42
     annihil
    0.40
    ic
    0.40
    ﺭ
    0.40
    ISC
    0.39
    ist
    0.38
    がお
    0.38
    ing
    0.36
     separators
    0.36
    POSITIVE LOGITS
    zelfde
    0.86
    ه
    0.73
    ン
    0.65
     consequently
    0.59
    lardan
    0.59
     сегодняшний
    0.58
    티브
    0.57
     поскольку
    0.57
    ları
    0.55
    это
    0.55
    Activations Density 0.078%

    No Known Activations