© 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. 45148
    Prev
    Next
    INDEX
    Explanations

    secret agents and operatives

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

    The neuron activates on words and phrases that denote spies or secret‐agent operatives (e.g., “agent,” “secret,” “covert,” “operative”).

    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
     annons
    -1.63
     perles
    -1.61
    ＇
    -1.55
     befolk
    -1.53
     wuß
    -1.48
     тази
    -1.46
    鈥
    -1.45
     særlig
    -1.45
     varmt
    -1.45
    sukienka
    -1.45
    POSITIVE LOGITS
     of
    1.50
    6
    1.49
     finds
    1.48
    i
    1.46
     почти
    1.46
    也有
    1.41
    5
    1.41
    7
    1.38
    f
    1.38
    In
    1.34
    Activations Density 0.040%

    No Known Activations