© 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. 39522
    Prev
    Next
    INDEX
    Explanations

    military units and courts

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

    The neuron activates on military unit designations that pair an ordinal number with a branch or unit type (e.g., “36th Rescue Squadron,” “146th Air Support Operations Squadron,” “63rd Indian Infantry Brigade”).

    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
     stupendous
    -1.31
     mangled
    -1.29
    Its
    -1.27
    ごちそう
    -1.26
     免费
    -1.26
     satisfactory
    -1.26
     its
    -1.26
    瞜
    -1.25
     sufficiently
    -1.24
    する方法
    -1.24
    POSITIVE LOGITS
    <bos>
    1.63
     of
    1.62
     to
    1.47
     thofe
    1.42
    eachtet
    1.40
     jā
    1.38
    ända
    1.38
    thday
    1.38
    PointerException
    1.36
    estanding
    1.33
    Activations Density 0.046%

    No Known Activations