© 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
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 5509
    Prev
    Next
    INDEX
    Explanations

    The neuron is highly activated by tokens that signal list items, section headings, or enumerations (e.g. dashes, category counts, “categories,” “common,” “concern”), so it’s looking for outline‐style or list‐style structural markers.

    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
     sebagainya
    0.66
    などに
    0.63
    (){}
    0.62
     (!)
    0.61
     وغیرہ
    0.61
     &/
    0.61
     इत्यादी
    0.61
     etcétera
    0.61
     etc
    0.60
    などで
    0.59
    POSITIVE LOGITS
     namely
    3.06
     yaitu
    2.84
    namely
    2.78
     yakni
    2.77
     Namely
    2.63
     iaitu
    2.34
     nämlich
    2.33
     namelijk
    2.31
    那就是
    2.17
     viz
    1.84
    Activations Density 0.581%

    No Known Activations