© 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. 11993
    Prev
    Next
    INDEX
    Explanations

    numbers, commas, dollar signs, or dates

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

    This neuron activates on numeric tokens—especially numbers with decimal points—i.e. floating-point figures.

    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
    9
    0.82
    1
    0.73
    0.71
    2
    0.68
    5
    0.65
    6
    0.64
    0
    0.64
    8
    0.61
    3
    0.60
    4
    0.58
    POSITIVE LOGITS
    ἲ
    0.55
    ॕ
    0.51
    <unused40>
    0.49
    <unused24>
    0.49
    <unused22>
    0.48
    épendant
    0.47
    ድግዳ
    0.47
    ocking
    0.46
    irte
    0.46
    ন্‌
    0.46
    Activations Density 0.170%

    No Known Activations