© 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. 6776
    Prev
    Next
    INDEX
    Explanations

    entity/attribute

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

    The neuron fires on tokens that are punctuation-based separators—especially forward slashes ("/") or hyphens—marking compound or category-style word joins.

    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
     связан
    0.82
     diphenyl
    0.75
     такой
    0.75
     মতো
    0.72
    这种
    0.70
    這種
    0.70
    ຕ້ອງ
    0.68
    
    0.68
     whose
    0.66
     hexagons
    0.66
    POSITIVE LOGITS
    初心者
    0.84
     सुद्धा
    0.80
    🧺
    0.80
    彡
    0.78
    _{
    0.76
    اح
    0.76
    _{\
    0.73
     Jullie
    0.73
    F
    0.72
    ım
    0.71
    Activations Density 0.759%

    No Known Activations