© 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. 31-GEMMASCOPE-2-RES-16K
    7. 10755
    Prev
    Next
    INDEX
    Explanations

    explicit content warning

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

    The neuron fires strongly on standalone metadata or heading‐style tokens—things like section labels, tags/categories, and other isolated domain-specific terms.

    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_31_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
    的
    0.65
    8
    0.53
    วิ
    0.51
    9
    0.50
    nya
    0.50
    の
    0.50
    ..
    0.50
    G
    0.50
    م
    0.50
    <eos>
    0.48
    POSITIVE LOGITS
     बढ़
    0.65
     QText
    0.63
    䥑
    0.63
    🚻
    0.63
    🔡
    0.62
     न्यूज
    0.62
    timevals
    0.61
    اداس
    0.60
     NONCLUSTERED
    0.60
     לאחר
    0.59
    Activations Density 1.342%

    No Known Activations