© 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. 3604
    Prev
    Next
    INDEX
    Explanations

    "able to", "generated solutions", "Enhancement model", "tuba toothpaste"

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

    The neuron is primarily detecting floating-point number tokens (i.e. decimal numbers).

    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
     ৫৭
    0.72
     emphatic
    0.64
     ৫৫
    0.61
     markedly
    0.60
     進
    0.58
     総
    0.58
     いき
    0.58
     била
    0.58
     chiefly
    0.57
     يق
    0.56
    POSITIVE LOGITS
     microfiber
    0.57
     स्क्रब
    0.52
     acrylic
    0.51
     enamel
    0.51
     gamme
    0.51
    rello
    0.51
    织
    0.51
    uyang
    0.50
    platte
    0.49
     hollywood
    0.49
    Activations Density 0.011%

    No Known Activations