© 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. 73854
    Prev
    Next
    INDEX
    Explanations

    legal or financial documents

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

    This neuron fires on section‐header or label tokens written in all caps (often with underscores or punctuation), such as constant names, docket headings, or table‐column titles.

    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
    h
    -3.03
    s
    -3.02
    ing
    -2.88
     bunda
    -2.80
     itſelf
    -2.77
    ”，
    -2.69
    e
    -2.67
     beb
    -2.61
    💓
    -2.52
     pourra
    -2.50
    POSITIVE LOGITS
    {
    3.58
    笑笑
    2.56
     \
    2.52
    中午
    2.44
     Seventy
    2.42
    1
    2.41
     Sixty
    2.33
     Historically
    2.31
     Naturally
    2.31
    on
    2.31
    Activations Density 0.007%

    No Known Activations