© 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. 128906
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on unusually long or rare tokens (e.g. “example,” “sale,” “config,” “API,” “WP”), effectively picking out infrequent or domain-specific words.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    software examples API

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    事が
    -1.18
    つづく
    -1.17
    ためには
    -1.11
    お楽しみに
    -1.09
    $}}
    -1.09
     desgaste
    -1.09
    んですよね
    -1.07
    事は
    -1.05
    んですね
    -1.02
    nancy
    -1.01
    POSITIVE LOGITS
     had
    1.04
    もちゃ
    0.98
    があった
    0.98
    ですが
    0.98
    それは
    0.98
    Makanan
    0.97
    استفاده
    0.96
    تحميل
    0.95
    がありますが
    0.95
     trattano
    0.95
    Activations Density 0.001%

    No Known Activations