© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 7639
    Prev
    Next
    INDEX
    Explanations

    custom exceptions" or "static variable"

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

    The neuron fires strongly on short functional or label‐type tokens used in headings, UI prompts, or section markers (e.g. “pic,” “Part,” “No,” “to,” “using,” “up,” “soon,” “is,” “it”).

    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
    ڑک
    0.77
     长
    0.75
    แหน่ง
    0.74
    جی
    0.74
    encils
    0.74
    فی
    0.72
    ائر
    0.71
     นับ
    0.71
     いう
    0.71
    میٹ
    0.70
    POSITIVE LOGITS
     zahlreiche
    0.76
    ➥
    0.76
     раствор
    0.75
    durch
    0.73
     различные
    0.73
    줌
    0.73
     сохран
    0.72
     abund
    0.72
     RECOMM
    0.72
     zdroj
    0.72
    Activations Density 0.001%

    No Known Activations