© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 15582
    Prev
    Next
    INDEX
    Explanations

    start of sentences or lists

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

    This neuron fires on the first word of a post or section heading—that is, the initial token (often a capitalized word) that marks the start of a message or heading.

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
    了一個
    -1.77
    着一个
    -1.41
    了一种
    -1.40
     ก็
    -1.36
    自行车
    -1.34
    使って
    -1.34
    了些
    -1.33
    两年
    -1.32
    lija
    -1.32
    遊んで
    -1.30
    POSITIVE LOGITS
    </b>
    1.69
    NECTIONS
    1.40
    ки
    1.40
     melap
    1.35
     []:
    1.35
    gados
    1.31
    >();
    1.27
    ȃ
    1.27
     menik
    1.27
    微博
    1.27
    Activations Density 0.298%

    No Known Activations