© 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. 127702
    Prev
    Next
    INDEX
    Explanations

    US military engagement abroad

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

    The neuron activates on mentions of U.S. or American organizations and their foreign‐policy actions (e.g. “American Friends of the Middle East,” “U.S. support,” “U.S. special forces,” “U.S. engagement”).

    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
     or
    -1.85
     our
    -1.73
     by
    -1.48
     some
    -1.32
     any
    -1.17
    ="
    -1.17
    )
    -1.16
     we
    -1.15
     five
    -1.13
     unserer
    -1.13
    POSITIVE LOGITS
    seitige
    1.28
    '[
    1.25
    ditionally
    1.25
     persiste
    1.23
    踉
    1.20
    sending
    1.18
    athered
    1.16
    sistance
    1.16
     supprimer
    1.14
    angebot
    1.14
    Activations Density 0.103%

    No Known Activations