© 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. 36153
    Prev
    Next
    INDEX
    Explanations

    direct statements and clarifications

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

    The neuron fires on common colloquial introductory or emphatic discourse markers (e.g. “let’s get something straight first,” “let it be said,” “I’m gonna say that a second time”).

    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
     such
    -0.98
     of
    -0.98
     closely
    -0.94
     leuke
    -0.93
     encontra
    -0.93
     све
    -0.90
    ₊
    -0.90
    そのまま
    -0.88
    自体は
    -0.88
     zoals
    -0.87
    POSITIVE LOGITS
     here
    1.38
    STATEMENT
    1.16
     bahwa
    1.08
     aquellos
    1.02
     عرضه
    1.00
    Therefore
    0.99
    Honestly
    0.98
     rằng
    0.98
    📢
    0.96
    ATORY
    0.95
    Activations Density 0.033%

    No Known Activations