© 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. 10-GEMMASCOPE-RES-131K
    4. 130694
    Prev
    Next
    INDEX
    Explanations

    idiomatic speech markers

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

    This neuron detects idiomatic or parenthetical “hedging” phrases—e.g. conversational qualifiers like “as it were,” “so to speak,” “as if,” or “as they say.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     SEÑ
    -1.72
    ſſen
    -1.71
    蛩
    -1.67
    琑
    -1.57
    tissimo
    -1.55
    bbling
    -1.52
    轫
    -1.45
     pokud
    -1.45
    铤
    -1.44
     Pasir
    -1.42
    POSITIVE LOGITS
     A
    1.57
    狃
    1.53
     sơn
    1.49
     jedoch
    1.48
     而
    1.46
     öst
    1.46
    
    1.45
    WER
    1.45
     Lots
    1.45
    liegenden
    1.43
    Activations Density 0.003%

    No Known Activations