© 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. 8834
    Prev
    Next
    INDEX
    Explanations

    where followed by descriptive clause

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

    The neuron fires strongly on infrequent, content‐bearing tokens—especially rare proper nouns or brand names.

    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
     while
    0.99
     mentre
    0.99
     when
    0.96
     когато
    0.85
    while
    0.84
     whereas
    0.82
     जब
    0.81
    당
    0.81
    when
    0.80
     enquanto
    0.80
    POSITIVE LOGITS
    soever
    1.02
     presumably
    0.98
    你可以
    0.92
     incidentally
    0.89
     dwells
    0.88
     admittedly
    0.81
     unscrupulous
    0.81
     ironically
    0.80
     lze
    0.79
     galima
    0.77
    Activations Density 0.139%

    No Known Activations