© 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. 24-GEMMASCOPE-2-RES-16K
    7. 9048
    Prev
    Next
    INDEX
    Explanations

    attributed to Charles M

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

    The neuron is highly activated by content‐bearing tokens—nouns, verbs, adjectives, numbers, and proper names—rather than by common function 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
     unstoppable
    0.53
     unfl
    0.50
    0.50
     explosive
    0.50
     desal
    0.49
    ार
    0.49
     одино
    0.49
    ஷ
    0.49
     damped
    0.48
     pessimistic
    0.48
    POSITIVE LOGITS
     Schrift
    0.72
     Ceux
    0.69
     ندرس
    0.68
     आवश्यकताओं
    0.68
     particulières
    0.68
    徬
    0.68
    ètres
    0.67
    mateurs
    0.67
    준다
    0.66
     ลักษณะ
    0.66
    Activations Density 0.001%

    No Known Activations