© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 71573
    Prev
    Next
    INDEX
    Explanations

    German, French, foreign language

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

    The neuron fires on specialized scientific or technical terms and titles—words that indicate professional roles (like physicist, inventor, surgeon) or domain‐specific concepts (like thermodynamics, neuro-, cryobiology, ligate) in scholarly or biographical 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_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
     those
    -1.78
     this
    -1.72
     here
    -1.67
     your
    -1.63
     will
    -1.56
     it
    -1.55
     each
    -1.48
     currently
    -1.45
    meneu
    -1.45
     (
    -1.41
    POSITIVE LOGITS
     religieuses
    1.78
     asisten
    1.68
     chré
    1.63
     burbu
    1.51
     sogar
    1.48
    暧
    1.47
     praktis
    1.47
     растений
    1.46
     judíos
    1.41
     religiosos
    1.38
    Activations Density 0.201%

    No Known Activations