© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 3202
    Prev
    Next
    INDEX
    Explanations

    dash separators after list items

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

    The neuron is keyed to the appearance of quantification—it lights up on words immediately surrounding numerical values (e.g. headers, labels, or items paired with a number).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    Looking at the high activations, I can see extremely large values (in the thousands) for tokens like "August" (4270), "December" (2161), "topics" (3829), "media" (3652), "play" (3612), "things" (

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    capitalized, title-like tokens that denote named entities, events/institutions, or time markers within lists and reports.

    oai_token-act-pair · gpt-5Triggered 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
    rollerskates
    0.61
    carouselExample
    0.55
    fromRGB
    0.54
    (",")
    0.54
     folgender
    0.54
    தாகும்
    0.52
    দেখিতে
    0.52
    parvec
    0.51
    browserAction
    0.50
    කය
    0.49
    POSITIVE LOGITS
     -
    4.50
     –
    4.28
     —
    3.81
     −
    3.47
     $-$
    3.38
    —
    3.34
     --
    3.28
     -"
    3.20
     -.
    3.17
     －
    3.14
    Activations Density 1.591%

    No Known Activations