© 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. 31-GEMMASCOPE-2-RES-16K
    7. 1826
    Prev
    Next
    INDEX
    Explanations

    The neuron detects numeric tokens—numbers and numeric values (e.g. float scores, list indices, ratings) in the text.

    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
     snug
    0.69
     handy
    0.68
     splendid
    0.68
     :-)
    0.68
     Helps
    0.66
     helps
    0.65
     🙂
    0.64
     versatile
    0.64
     comfy
    0.63
     rejuvenating
    0.63
    POSITIVE LOGITS
    😡
    1.70
    🤬
    1.69
    😠
    1.54
     appalling
    1.52
     disgusting
    1.50
    👎
    1.46
     disgraceful
    1.45
     deplorable
    1.42
    🤮
    1.39
     unprofessional
    1.38
    Activations Density 0.478%

    No Known Activations