© 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. 73062
    Prev
    Next
    INDEX
    Explanations

    items correspond with respectively

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

    The neuron is picking out domain-specific or “named” tokens – especially abbreviations, acronyms, and other technical identifiers (e.g. uppercase codes, parenthesized terms, quoted/percent-containing strings).

    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
     здесь
    -0.84
     bire
    -0.83
    diol
    -0.79
    Zheng
    -0.79
    liš
    -0.79
    mbang
    -0.78
    bangan
    -0.76
    Dispatcher
    -0.76
     aquí
    -0.75
     ibang
    -0.75
    POSITIVE LOGITS
     respectively
    3.41
    respectively
    2.47
     respectivamente
    2.41
     respectivement
    2.22
     соответственно
    1.95
    それぞれ
    1.77
     respective
    1.77
     rispet
    1.68
     respec
    1.55
    分别
    1.45
    Activations Density 0.163%

    No Known Activations