© 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. 3333
    Prev
    Next
    INDEX
    Explanations

    so-called anatomical terms

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

    The neuron fires on semantically “heavy” content words—key nouns, verbs, and prepositions that carry the core meaning of a technical or descriptive sentence—rather than on light function words.

    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
     对象
    0.65
     Hiện
    0.61
    otrex
    0.60
     例
    0.60
     Trên
    0.60
     高校
    0.60
     など
    0.59
    và
    0.59
    ắ
    0.59
    illar
    0.59
    POSITIVE LOGITS
     cosidd
    1.17
     sogenannten
    1.02
     sogenannte
    1.02
     tzw
    0.98
     tzv
    0.98
     socalled
    0.94
     sogen
    0.77
     ruthless
    0.74
     zog
    0.74
     nahezu
    0.67
    Activations Density 0.198%

    No Known Activations