© 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. 31-GEMMASCOPE-2-RES-16K
    7. 8469
    Prev
    Next
    INDEX
    Explanations

    past tense verb suffix

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

    This neuron strongly activates on non-Latin (CJK) characters, i.e. Japanese or Korean text.

    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
    들
    0.89
    인데
    0.68
    쪽
    0.67
    이다
    0.64
    가지
    0.63
    వ్వు
    0.63
    에요
    0.62
    까
    0.62
    들이
    0.61
    곳
    0.61
    POSITIVE LOGITS
    した
    1.27
    したのは
    1.27
    された
    1.22
    され
    1.16
    される
    1.15
    しました
    1.13
    されました
    1.13
    されて
    1.12
    する
    1.05
    すると
    1.03
    Activations Density 0.015%

    No Known Activations