© 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. 22-GEMMASCOPE-RES-131K
    7. 67800
    Prev
    Next
    INDEX
    Explanations

    La Chaise, Say Yes, Samsung, Cholesky decomposition

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

    The neuron fires on mathematical matrix expressions and operations—e.g. symbols like det(·), X, Δx, inverses, eigen decompositions—i.e. linear‐algebra notation.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     höl
    -1.36
    整个人
    -1.12
    öffel
    -1.09
     hivat
    -1.07
    pendientes
    -1.07
    Concepto
    -1.05
     euró
    -1.05
    ophobic
    -1.05
     esetben
    -1.05
    摇了摇头
    -1.04
    POSITIVE LOGITS
     like
    1.08
    Ϣ
    1.04
    des
    1.04
    erobic
    1.03
     what
    0.99
    ométrie
    0.98
    wsze
    0.95
     some
    0.94
     has
    0.93
     consisting
    0.93
    Activations Density 0.005%

    No Known Activations