© 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. 31-GEMMASCOPE-2-RES-16K
    7. 1533
    Prev
    Next
    INDEX
    Explanations

    column separators

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

    This neuron detects the “Q” question‐label tokens (and their accompanying numeric/dot punctuation) that mark question 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_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
     rq
    0.57
     след
    0.55
     kw
    0.53
    ět
    0.52
    死去
    0.52
     generalizations
    0.50
    これにより
    0.50
    給
    0.50
    眞
    0.50
    емых
    0.49
    POSITIVE LOGITS
     thing
    0.71
    şey
    0.69
     şey
    0.65
    thing
    0.62
    dib
    0.60
    yers
    0.59
    yer
    0.57
     tablespoon
    0.57
    ose
    0.55
    memb
    0.54
    Activations Density 0.026%

    No Known Activations