© 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. 7348
    Prev
    Next
    INDEX
    Explanations

    breakdown or explanation

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

    The neuron is acting like a TF–IDF filter: it lights up on content-bearing, relatively rare words (proper names, descriptive adjectives/verbs) rather than on high-frequency 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
     stupid
    0.92
     моей
    0.85
    私が
    0.82
    所以我
    0.82
     maravilh
    0.79
     sooo
    0.77
    在我们
    0.77
     wonderful
    0.76
     ДО
    0.76
     soooo
    0.75
    POSITIVE LOGITS
    球迷
    0.81
     선수
    0.73
    ESPN
    0.73
     ملعب
    0.72
    ক্ষেপ
    0.68
     കീഴ
    0.68
     Onun
    0.68
     തന്റെ
    0.67
    ángulos
    0.67
    staben
    0.67
    Activations Density 0.477%

    No Known Activations