© 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. 24-GEMMASCOPE-2-RES-16K
    7. 1045
    Prev
    Next
    INDEX
    Explanations

    parenthetical citations

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

    The neuron strongly activates on numeric year tokens (publication dates) in citations or 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_24_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
     the
    0.68
     konci
    0.61
     one
    0.59
     this
    0.57
    0.57
     =
    0.57
     là
    0.56
     snatched
    0.56
     satu
    0.56
     chopped
    0.55
    POSITIVE LOGITS
    arxiv
    0.60
     Additionally
    0.55
     Similarly
    0.55
    reviewed
    0.54
     Preprint
    0.52
    Furthermore
    0.52
    ῴ
    0.52
    論文
    0.52
    ].
    0.51
    Interestingly
    0.51
    Activations Density 0.024%

    No Known Activations