© 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. 31-GEMMASCOPE-2-RES-16K
    7. 12984
    Prev
    Next
    INDEX
    Explanations

    code with numbers and punctuation

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

    This neuron strongly activates on numeric tokens and the surrounding punctuation typical of bibliographic citations (years, volume and page numbers in parentheses).

    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
     canonical
    0.54
     തുടങ്ങി
    0.52
    ٿ
    0.51
     antecedent
    0.49
    특별시
    0.49
     canceled
    0.49
     চেয়েছিলেন
    0.49
    جهات
    0.48
     ทด
    0.48
     द्वार
    0.47
    POSITIVE LOGITS
     n
    2.03
    ｎ
    1.47
    n
    1.36
     nd
    1.30
    nYou
    1.14
    nThe
    1.12
     н
    1.10
    𝑛
    1.10
     nCurrent
    1.07
    nd
    1.05
    Activations Density 0.094%

    No Known Activations