© 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. 12685
    Prev
    Next
    INDEX
    Explanations

    closing quote punctuation before said

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

    This neuron identifies the quotation‐mark tokens that signal direct speech or quoted passages.

    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
    だけではなく
    -1.04
    ॄ
    -1.03
     offizielle
    -1.02
    ሺ
    -1.00
    isure
    -1.00
    ennom
    -1.00
    をつけて
    -0.96
    essment
    -0.96
    ortheast
    -0.96
    okkal
    -0.95
    POSITIVE LOGITS
     olarak
    1.06
     added
    1.05
    ™
    1.02
     remainder
    1.02
     it
    1.01
     afirma
    1.00
     again
    0.95
    zé
    0.94
    ®
    0.91
    还能
    0.90
    Activations Density 0.005%

    No Known Activations