© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 15826
    Prev
    Next
    INDEX
    Explanations

    python code or script

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

    This neuron activates strongly on uncommon or domain-specific tokens—proper nouns, technical jargon, acronyms, and other low-frequency “special” words.

    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
    aság
    0.98
    습니다
    0.93
    ్‌
    0.92
     TRANSPORTURI
    0.92
    től
    0.88
     Foods
    0.88
     numérique
    0.88
    많
    0.88
    ovaní
    0.86
     požad
    0.86
    POSITIVE LOGITS
    he
    0.97
    ischer
    0.95
    -
    0.91
    atively
    0.89
    :
    0.89
    طرف
    0.89
    ভাবে
    0.86
    ure
    0.85
     chime
    0.85
    `
    0.85
    Activations Density 8.030%

    No Known Activations