© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 100407
    Prev
    Next
    INDEX
    Explanations

    starting clause

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

    This neuron fires on common discourse markers and short function words that kick off spoken turns or dialogue lines (e.g. “I,” “And,” “Well,” “You,” “Hi,” “Can,” “Please”).

    oai_token-act-pair · o4-miniTriggered 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
    Not in Any Lists

    No Comments

    Negative Logits
    ↵
    -1.84
    8
    -1.82
    "
    -1.82
    $
    -1.78
    Don
    -1.58
    6
    -1.52
     自分
    -1.52
    m
    -1.52
    5
    -1.52
    眼睛
    -1.48
    POSITIVE LOGITS
     suivent
    1.80
     murky
    1.59
    ;;)
    1.56
    👿
    1.55
     :{
    1.53
     commencent
    1.53
    ׅ
    1.52
    虓
    1.48
     --;
    1.48
     prized
    1.47
    Activations Density 0.015%

    No Known Activations