© 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. 4837
    Prev
    Next
    INDEX
    Explanations

    urgency, help, accurate

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

    The neuron fires strongly on the very first content token of a new section or paragraph, i.e. it detects “start‐of‐segment” 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
     This
    0.65
    This
    0.64
     这
    0.63
     બીજી
    0.62
     The
    0.61
     इस्तेमाल
    0.60
     Chimie
    0.59
    स्तिष्क
    0.59
     폭
    0.59
     왼쪽
    0.58
    POSITIVE LOGITS
    ↵↵
    0.92
     underserved
    0.60
    agricult
    0.59
    ↵↵↵
    0.57
     southernmost
    0.57
    Вер
    0.56
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    0.55
    峇
    0.55
     showcases
    0.55
    দল
    0.55
    Activations Density 0.041%

    No Known Activations