© 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. 22-GEMMASCOPE-RES-131K
    7. 15003
    Prev
    Next
    INDEX
    Explanations

    Qatar and Venice

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

    The main thing this neuron does is detect mentions of “Qatar” (including its adjectival form).

    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
    k
    -2.19
    начально
    -2.11
    ro
    -1.98
    ed
    -1.93
     =
    -1.93
    d
    -1.89
    os
    -1.89
    ing
    -1.88
    i
    -1.88
    ch
    -1.86
    POSITIVE LOGITS
    ’
    2.09
    鹀
    2.03
    赑
    1.94
    震惊
    1.80
    
    1.71
    〮
    1.71
    らしいです
    1.70
    쏙
    1.66
    禵
    1.66
     brasileiros
    1.63
    Activations Density 0.035%

    No Known Activations