© 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. 7796
    Prev
    Next
    INDEX
    Explanations

    core idea or function

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

    The neuron fires strongly on informative content words—especially significant nouns, verbs, and adjectives carrying the main topic (i.e. “wedding,” “elaborating,” “program,” “challenge,” etc.)—and ignores 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
     혹은
    0.75
     或者
    0.73
     மற்றும்
    0.72
     atau
    0.71
     หรือ
    0.70
    或者
    0.70
     or
    0.68
     hoặc
    0.68
    ませんので
    0.66
     eller
    0.66
    POSITIVE LOGITS
    結論
    0.80
    ។
    0.71
    。.
    0.70
    思路
    0.69
     sistemat
    0.68
    今日の
    0.68
    最初の
    0.67
    下一步
    0.66
    用い
    0.66
    ようやく
    0.66
    Activations Density 0.090%

    No Known Activations