© 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
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 2143
    Prev
    Next
    INDEX
    Explanations

    names, dates, and code keywords

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

    The neuron detects punctuation tokens in source code, especially dots in fully-qualified names and semicolons ending statements.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     worth
    -0.87
     had
    -0.86
     trying
    -0.84
     despite
    -0.84
     during
    -0.82
     seeking
    -0.81
     making
    -0.77
    chae
    -0.77
     want
    -0.76
    cą
    -0.76
    POSITIVE LOGITS
    argli
    0.99
     regulador
    0.86
     clouded
    0.85
     phare
    0.84
     Bangkok
    0.83
     nisk
    0.83
    när
    0.82
     materiaal
    0.81
     veneno
    0.81
    DESIGN
    0.81
    Activations Density 0.003%

    No Known Activations