© 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. 65616
    Prev
    Next
    INDEX
    Explanations

    math and programming terms

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

    The neuron activates on uncommon, domain-specific technical or scientific terms (i.e. rare jargon tokens).

    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
    क्या
    -1.08
     millon
    -1.05
    ielles
    -1.01
     first
    -1.00
    obiettivo
    -0.96
    ácie
    -0.94
    
    -0.92
     viaj
    -0.91
     during
    -0.91
     whose
    -0.90
    POSITIVE LOGITS
    を読む
    0.97
    すっ
    0.94
    を読
    0.93
    Hmmm
    0.93
    びっくり
    0.93
     trám
    0.91
    Ahh
    0.90
     krijgt
    0.90
     tragé
    0.90
     incrível
    0.89
    Activations Density 0.401%

    No Known Activations