© 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
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 104877
    Prev
    Next
    INDEX
    Explanations

    code comment or block

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

    horizontal line separators made of repeated dashes or other characters.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron activates on long runs of hyphens (divider or separator lines).

    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
    ↵↵
    -3.00
    翃
    -2.92
     zahlreichen
    -2.91
    骉
    -2.88
    o
    -2.73
     Lately
    -2.52
     Somewhere
    -2.50
     Choosing
    -2.45
    )
    -2.45
     ใน
    -2.41
    POSITIVE LOGITS
    寕
    3.63
    嗞
    3.58
    ꠥ
    3.47
    亖
    2.95
    祘
    2.95
    s
    2.92
    螄
    2.89
     عروس
    2.81
    𖥦
    2.73
    镦
    2.70
    Activations Density 0.004%

    No Known Activations