© 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. 1539
    Prev
    Next
    INDEX
    Explanations

    Option 1

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

    This neuron selectively activates on numeric tokens (e.g. season/episode numbers, figure/claim numbers, dates, statistics, and other 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
    ብሰ
    0.75
    centerX
    0.73
    ຜະລິດຕ
    0.72
    hasOwnProperty
    0.65
    虐
    0.64
    낡
    0.64
    exempt
    0.64
    য়েছে
    0.63
    trunc
    0.63
    usable
    0.63
    POSITIVE LOGITS
     III
    1.96
     VII
    1.77
     II
    1.72
     №
    1.66
     VIII
    1.64
     XII
    1.57
     XIII
    1.55
     IV
    1.47
     nº
    1.46
     iii
    1.43
    Activations Density 0.903%

    No Known Activations