© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 91760
    Prev
    Next
    INDEX
    Explanations

    phrases starting with common words or punctuation

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

    This neuron fires on document‐structure tokens—e.g. uppercase headers, speaker labels, parentheses, “BEGIN/END CLIP,” page markers, timestamps—i.e. formatting and transcript metadata rather than normal text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     บาง
    -0.80
     まとめ
    -0.78
     ilang
    -0.76
     Học
    -0.75
     Reels
    -0.75
    ឋ
    -0.73
     Appropriations
    -0.70
    ᱼ
    -0.69
    經歷
    -0.69
     responded
    -0.68
    POSITIVE LOGITS
    緘
    0.86
    serializers
    0.75
     memutus
    0.74
    جوايز
    0.73
    AddComponent
    0.73
    pandemic
    0.71
     наслед
    0.71
     Tonight
    0.71
     choose
    0.70
     सम
    0.70
    Activations Density 0.009%

    No Known Activations