© 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. 34-GEMMASCOPE-RES-131K
    4. 45220
    Prev
    Next
    INDEX
    Explanations

    technical system software program

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

    The neuron activates on high‐frequency “function” or stop words (articles, conjunctions, question words, etc.) rather than on substantive content words.

    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
     ligeramente
    -0.93
    これだけ
    -0.82
    utiliz
    -0.82
    cciones
    -0.79
    oluciones
    -0.79
    按键
    -0.77
    yeron
    -0.76
    quickly
    -0.76
     ceci
    -0.76
     Martínez
    -0.76
    POSITIVE LOGITS
     system
    3.50
     software
    2.47
    系统
    2.44
     program
    2.19
     algorithm
    2.17
    システム
    2.16
     система
    2.14
    system
    2.00
     sistema
    1.98
    系統
    1.91
    Activations Density 0.227%

    No Known Activations