© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 431
    Prev
    Next
    INDEX
    Explanations

    higher or lower than

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

    The neuron fires on single tokens that convey a comparative or quality boost—e.g. “better,” “higher,” “excellent,” “great”—so it’s detecting words that signal improved performance or superior quality.

    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
    ITION
    0.52
     КАК
    0.52
    IVING
    0.51
    untitled
    0.49
    ITATE
    0.48
    𝓸
    0.48
     пару
    0.47
    ()'
    0.47
    었어요
    0.47
    ריך
    0.46
    POSITIVE LOGITS
     compared
    0.61
     (>
    0.61
     exceed
    0.56
    而且
    0.56
    पेक्षा
    0.55
     exceeds
    0.55
    高于
    0.54
     whereas
    0.52
     than
    0.51
     ஏற
    0.50
    Activations Density 0.000%

    No Known Activations