© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 3430
    Prev
    Next
    INDEX
    Explanations

    audience or attendees

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

    The neuron detects small‐quantity expressions and numeric quantifiers (e.g. “one,” “only,” “a handful,” “most of”) indicating limited group sizes.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     lived
    0.64
     mutation
    0.63
     duo
    0.63
     tööt
    0.61
     akut
    0.59
     mutated
    0.58
    做事
    0.57
     Duo
    0.57
    饵
    0.55
     mutations
    0.54
    POSITIVE LOGITS
     audience
    2.13
     crowd
    2.05
    audience
    2.03
     Audience
    1.90
    观众
    1.76
     crowds
    1.75
    觀眾
    1.67
     Crowd
    1.66
     audiences
    1.62
     applause
    1.58
    Activations Density 0.535%

    No Known Activations