© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 93048
    Prev
    Next
    INDEX
    Explanations

    code, scripts, libraries

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

    The neuron fires on code‐markup tokens—especially HTML tags (like “<”, “script”, “src”) and JavaScript file references (“.js”).

    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
     estreno
    -1.25
    ₱
    -1.23
    者に
    -1.17
     lentejuelas
    -1.16
    [];
    
    -1.12
    Vorlage
    -1.08
    ₽
    -1.08
     acabados
    -1.05
     family
    -1.05
    )";
    -1.05
    POSITIVE LOGITS
     najdete
    1.07
    Sied
    1.02
    Juri
    1.01
    Svet
    1.01
     and
    1.01
    Mó
    0.98
    Probl
    0.96
     ร์
    0.95
    Dota
    0.94
    стойчи
    0.94
    Activations Density 0.003%

    No Known Activations