© 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. 127692
    Prev
    Next
    INDEX
    Explanations

    symbols followed by names or titles

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

    The neuron responds to document‐level metadata and formatting markers—like byline/timestamp labels, signature tildes, “Guest Post:” headers, “Submitted by,” “Posted” dates, etc.—rather than regular content words.

    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
    ):
    -1.42
    ↵↵
    -1.36
     amerikanischen
    -1.30
     Initially
    -1.23
     -
    -1.21
     F
    -1.21
     weiterhin
    -1.21
     wirtschaft
    -1.20
     folgenden
    -1.19
     K
    -1.18
    POSITIVE LOGITS
     about
    1.86
     geene
    1.77
     hals
    1.73
     stoff
    1.73
     daer
    1.73
     ijs
    1.67
     staffel
    1.65
    蛱
    1.65
     damen
    1.64
     wata
    1.62
    Activations Density 0.016%

    No Known Activations