© 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. 6320
    Prev
    Next
    INDEX
    Explanations

    medical and health conditions

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

    The main thing this neuron does is detect explicit numeric values and unit‐style measurements (e.g. step numbers, temperatures, quantities) in instructional or specification 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
    中国
    0.78
    涵盖
    0.73
     中国
    0.70
    一一
    0.68
    创意
    0.65
    全国
    0.62
    美国
    0.61
    🏆
    0.61
    未来
    0.61
    国
    0.60
    POSITIVE LOGITS
     leakage
    0.68
     uneas
    0.66
     각종
    0.65
     phenomena
    0.64
     sparsim
    0.64
     заболевания
    0.64
     deformation
    0.63
    autions
    0.63
     Quar
    0.63
     phénomène
    0.63
    Activations Density 0.121%

    No Known Activations