© 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. 8950
    Prev
    Next
    INDEX
    Explanations

    code snippets and commands

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

    The neuron fires on parts of academic reference strings—especially venue names (e.g. “Pattern Recognition,” “CV,” “PR,” “ICPR”) and the adjoining volume, issue, year, and page‐number 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
    constexpr
    0.73
     constexpr
    0.66
     empowers
    0.65
     microbiome
    0.65
    Lett
    0.60
     takeaway
    0.60
     }%
    0.59
     CRISPR
    0.59
     profinite
    0.58
    स्म
    0.57
    POSITIVE LOGITS
     cms
    0.66
    病人
    0.57
    swf
    0.57
    argc
    0.56
     basename
    0.55
     жел
    0.54
    笨
    0.54
    尺寸
    0.54
     कार्ट
    0.54
     команда
    0.54
    Activations Density 0.007%

    No Known Activations