© 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. 12066
    Prev
    Next
    INDEX
    Explanations

    `m` `v` `qs` `self` `Rating`

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

    The neuron fires most strongly on standalone numeric tokens—especially multi‐digit numbers (e.g. years or other long numerals).

    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
    e
    0.88
    /
    0.84
    Reading
    0.78
    Jeremy
    0.77
    </em>
    0.75
    During
    0.72
    Between
    0.72
    요
    0.71
    <u>
    0.71
    eee
    0.70
    POSITIVE LOGITS
     isomers
    0.85
     theorems
    0.82
     PBMCs
    0.82
     enshrined
    0.81
     opined
    0.80
     isoforms
    0.77
     nanofibers
    0.77
     partisans
    0.77
     فلسط
    0.76
     molyb
    0.75
    Activations Density 0.000%

    No Known Activations