© 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. 25506
    Prev
    Next
    INDEX
    Explanations

    air conditioning, curiosity phrases, people, android, query, soap, stocks, market

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

    This neuron detects numeric tokens—especially list or sequence indices and other standalone numbers.

    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
    を重ね
    -1.09
     mikina
    -0.99
     obuv
    -0.98
    ꈊ
    -0.95
     verra
    -0.95
    schrank
    -0.93
    ویزی
    -0.91
     juzg
    -0.90
     Escribe
    -0.90
     Vaz
    -0.88
    POSITIVE LOGITS
    uo
    0.96
    â
    0.91
    or
    0.89
    </em>
    0.88
     эти
    0.88
    en
    0.86
    ub
    0.86
    u
    0.85
    r
    0.85
    尤其是
    0.85
    Activations Density 0.049%

    No Known Activations