© 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. 13835
    Prev
    Next
    INDEX
    Explanations

    line items

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

    This neuron fires on the very first real word after a new segment’s start (<bos>), i.e. it detects “beginning-of-text” tokens.

    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
    ިން
    0.70
    氻
    0.68
     infall
    0.68
    em
    0.67
    部落
    0.67
    当作
    0.67
    transmitted
    0.66
    izations
    0.66
    鮑
    0.66
    做一个
    0.65
    POSITIVE LOGITS
    на
    0.91
     üzer
    0.77
     erinn
    0.75
    طق
    0.74
    ق
    0.74
     cù
    0.73
    𝑰
    0.73
    ıp
    0.73
    ні
    0.72
     ciò
    0.71
    Activations Density 0.664%

    No Known Activations