© 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
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 8500
    Prev
    Next
    INDEX
    Explanations

    code blocks and ascii art

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

    The neuron strongly activates on long runs or clusters of punctuation characters (e.g. sequences of periods, dashes, and ellipses).

    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
    ós
    0.38
     —¿
    0.38
     |
    0.37
    📅
    0.37
    ól
    0.34
     "[
    0.34
    काल
    0.34
     Ara
    0.33
    ‼
    0.33
    Ord
    0.33
    POSITIVE LOGITS
    ::::
    0.45
    ::::::
    0.40
     মৃত্যু
    0.39
    )))))
    0.37
     www
    0.36
    ourke
    0.36
    wwww
    0.36
     সংখ্যাগরিষ্ঠ
    0.36
     ..........
    0.35
     AAA
    0.35
    Activations Density 0.033%

    No Known Activations