© 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. 22-GEMMASCOPE-RES-131K
    4. 12765
    Prev
    Next
    INDEX
    Explanations

    quoted speech or direct quotes

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

    The neuron detects the colon punctuation that introduces or precedes a direct quotation (the “said:” marker before quoted speech).

    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
    ["_
    -1.27
     povezave
    -1.22
    ネマ
    -1.18
    يتر
    -1.12
     remedio
    -1.12
    providedIn
    -1.09
    }],
    
    -1.09
     putea
    -1.09
    aná
    -1.08
    рк
    -1.06
    POSITIVE LOGITS
     referirse
    1.23
     praktisch
    1.20
     “
    1.18
     resaltar
    1.17
     technisch
    1.13
     heilige
    1.11
    creens
    1.09
    祛
    1.08
     größe
    1.06
    grazie
    1.06
    Activations Density 0.004%

    No Known Activations