© 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. 24-GEMMASCOPE-2-RES-16K
    7. 3348
    Prev
    Next
    INDEX
    Explanations

    numbers like 0, 1, 2, 3

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

    The neuron activates on numeric tokens—especially floating‐point or decimal numbers—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_24_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
     それ
    0.65
    awca
    0.59
     Algorithms
    0.59
     Entscheidung
    0.58
     fopen
    0.58
    ．
    0.58
     ausges
    0.57
    ق
    0.57
    ፡፡
    0.56
     overfitting
    0.56
    POSITIVE LOGITS
    хожу
    0.70
    dominated
    0.67
     darkened
    0.64
    ходили
    0.64
    driven
    0.61
    tried
    0.61
    мам
    0.60
    𝐩
    0.59
    য়েড
    0.59
     tainted
    0.59
    Activations Density 0.080%

    No Known Activations