© 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. 2075
    Prev
    Next
    INDEX
    Explanations

    history and time concepts

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

    This neuron detects isolated, capitalized “header” words or topic labels (single-word nouns like “Friends,” “History,” “Nature,” “God”) that stand alone as section or paragraph headings.

    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
     excerpts
    0.49
    0.48
     menus
    0.47
     galvanometer
    0.46
     landscape
    0.46
    :
    0.45
     landscapes
    0.45
    []
    0.45
     intensities
    0.43
     antik
    0.43
    POSITIVE LOGITS
     hath
    0.68
     doth
    0.62
    自有
    0.61
     Ừ
    0.60
     цьому
    0.59
     Knows
    0.59
     dritte
    0.59
     그거
    0.59
     seule
    0.58
     vuole
    0.57
    Activations Density 0.001%

    No Known Activations