© 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. 34099
    Prev
    Next
    INDEX
    Explanations

    title context

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

    The neuron fires strongly on template‐language delimiters (e.g. `{% … %}`, `<% … %>`, `extends`, `.html`), i.e. markers of server-side or client-side template syntax.

    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
     Tamaño
    -1.09
     herhangi
    -1.06
     Ně
    -1.05
     Secrétaire
    -1.02
    雨が
    -1.02
     Capitán
    -0.98
    bản
    -0.98
     Accesorios
    -0.97
     Bretaña
    -0.97
    ៚
    -0.97
    POSITIVE LOGITS
    他們
    1.08
    Instead
    1.07
     and
    1.05
    他们
    1.05
    They
    0.91
     chuy
    0.91
    我們
    0.90
     обычно
    0.90
    What
    0.89
    Vidi
    0.86
    Activations Density 0.007%

    No Known Activations