© 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. 15466
    Prev
    Next
    INDEX
    Explanations

    Multilingual programming code snippets

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

    The neuron is detecting markup delimiters, especially the angle‐bracket boundaries (“<” and “>”) that denote XML/HTML tags.

    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
     endregion
    -1.03
     this
    -1.03
     ـ
    -0.99
     pö
    -0.98
    awarkan
    -0.95
    vereignty
    -0.95
    sistema
    -0.94
    
    -0.91
     if
    -0.91
     our
    -0.91
    POSITIVE LOGITS
    N
    1.02
    T
    1.00
    L
    0.98
     intermitt
    0.98
    τούν
    0.96
    H
    0.96
    范文
    0.95
     ‘‘
    0.94
    феры
    0.94
    Törté
    0.92
    Activations Density 0.001%

    No Known Activations