© 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. 86907
    Prev
    Next
    INDEX
    Explanations

    idiomatic phrases

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

    The neuron fires strongly on tokens that are part of common English idioms or fixed figurative expressions (e.g. “same page,” “bite the bullet,” “short end of the stick,” “stick to your guns,” etc.).

    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
    為什麼
    -1.73
    during
    -1.58
    受欢迎
    -1.55
    .
    -1.54
    adalah
    -1.48
    rse
    -1.45
     granja
    -1.42
     Brasileira
    -1.40
     výrob
    -1.39
     muchas
    -1.36
    POSITIVE LOGITS
    ことも
    2.08
     of
    1.84
     PUSTAKA
    1.55
    man
    1.53
    ricanes
    1.52
    のも
    1.44
     требует
    1.44
    主流
    1.41
    </strong>
    1.39
     also
    1.39
    Activations Density 0.127%

    No Known Activations