© 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. 7491
    Prev
    Next
    INDEX
    Explanations

    This neuron relates to **functional, problem, method, information, relation, data, work, condition**

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

    The neuron flags foreign‐language tokens, especially those in non-Latin scripts (e.g. Cyrillic) or bearing diacritics.

    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
     різні
    0.84
     ένας
    0.81
     такі
    0.79
     एजेंस
    0.76
     инструменты
    0.74
     घटनाएं
    0.74
     nowe
    0.73
     tyto
    0.72
    ские
    0.71
     такие
    0.71
    POSITIVE LOGITS
    问题的
    1.04
    方法的
    1.04
    信息的
    1.01
    功能的
    1.00
    关系的
    0.96
    能力的
    0.95
    内容的
    0.95
    条件的
    0.93
    模型的
    0.93
    活动的
    0.93
    Activations Density 0.055%

    No Known Activations