© 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. 66332
    Prev
    Next
    INDEX
    Explanations

    code syntax and structure

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

    This neuron is keyed to code-syntax tokens—especially the dollar sign that introduces PHP variables (and similar code symbols and numeric literals).

    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
    zdroj
    -1.15
     škola
    -1.15
    ciclop
    -1.11
    camisetas
    -1.10
    spodnie
    -1.09
     påvir
    -1.09
     onderwerp
    -1.08
    ducir
    -1.08
     føl
    -1.07
     bestäm
    -1.07
    POSITIVE LOGITS
     geograf
    0.95
    M
    0.94
     religi
    0.93
    ber
    0.93
    </em>
    0.91
     درمانی
    0.88
    T
    0.88
     hecta
    0.88
    然后
    0.87
    mathrm
    0.87
    Activations Density 0.023%

    No Known Activations