© 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. 24-GEMMASCOPE-2-RES-16K
    7. 13280
    Prev
    Next
    INDEX
    Explanations

    Editor, audience, rules

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

    The neuron strongly activates on numeric tokens—especially decimal or floating-point numbers—wherever they appear 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
     umož
    0.90
     baar
    0.82
     VARIABLES
    0.81
     FBS
    0.81
    ﮏ
    0.80
     druż
    0.80
    ъем
    0.80
     fluido
    0.80
     paciente
    0.78
     MPH
    0.78
    POSITIVE LOGITS
    р
    0.80
    今日
    0.76
    说
    0.74
    azón
    0.73
    問題
    0.71
    ፈጥ
    0.71
     saying
    0.70
    ணா
    0.69
    ানা
    0.68
    原
    0.67
    Activations Density 0.000%

    No Known Activations