© 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. 22-GEMMASCOPE-RES-131K
    7. 9702
    Prev
    Next
    INDEX
    Explanations

    About Lock You a Filing What a ’ The Tagged

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

    The neuron detects the first (often capitalized) word of a document’s title or heading.

    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
    IFICA
    -0.97
    льня
    -0.97
     parlamentar
    -0.97
     idioma
    -0.94
    mole
    -0.93
    眞
    -0.92
    🍋
    -0.92
     Electoral
    -0.92
     galeria
    -0.91
    zenden
    -0.90
    POSITIVE LOGITS
    wantErr
    1.05
     verwende
    1.02
    Also
    0.98
    昨日の
    0.98
    ЧИТА
    0.97
     proposte
    0.97
    According
    0.96
     層
    0.92
    Lors
    0.91
    昨日は
    0.91
    Activations Density 0.001%

    No Known Activations