© 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. 34-GEMMASCOPE-RES-131K
    7. 82747
    Prev
    Next
    INDEX
    Explanations

    idiomatic phrases and metaphors

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

    The neuron is specialized for spotting multi-word idiomatic or metaphorical expressions.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    Wrap
    -0.88
     blandt
    -0.86
     afloat
    -0.82
    hésite
    -0.81
    Læs
    -0.81
     gennem
    -0.81
     daardoor
    -0.80
     tæt
    -0.78
     daarbij
    -0.77
    ーション
    -0.77
    POSITIVE LOGITS
     proverbial
    1.70
     metaphorical
    1.20
     figurative
    0.94
     many
    0.91
    difice
    0.87
    economic
    0.85
    meta
    0.84
    operability
    0.83
     vielen
    0.80
    onom
    0.77
    Activations Density 0.076%

    No Known Activations