© 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. 9383
    Prev
    Next
    INDEX
    Explanations

    never return

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

    The neuron strongly responds to names and titles—that is, proper nouns (especially song, soundtrack, or other title text) and other capitalized or named-entity tokens.

    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
    多元
    0.69
    硕
    0.65
    统计
    0.58
    迭代
    0.58
    整個
    0.58
    擎
    0.58
    亚洲
    0.57
    亚太
    0.57
    整个
    0.57
    忽略
    0.55
    POSITIVE LOGITS
     ♪
    1.06
    ♪
    0.97
     SNS
    0.90
     부탁
    0.89
    おすすめ
    0.88
     ◎
    0.86
     じゃ
    0.85
     お
    0.84
     생각
    0.83
     그런
    0.83
    Activations Density 0.189%

    No Known Activations