© 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. 71513
    Prev
    Next
    INDEX
    Explanations

    facial expressions and poses

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

    The neuron responds in proportion to token length—it fires strongest on unusually long words.

    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
    ּוֹ
    -1.55
    ֶּ
    -1.27
    ַּ
    -1.20
     السياس
    -1.17
    CLUSIVE
    -1.15
     zamiast
    -1.13
    ̵
    -1.12
    meldet
    -1.10
    omal
    -1.05
     ermöglichen
    -1.05
    POSITIVE LOGITS
    っております
    1.10
     游戏
    1.10
    “……”
    1.09
    mäh
    1.09
     Ольга
    1.07
     表情
    1.06
     expression
    1.05
    悸
    1.05
     安装
    1.05
    licante
    1.05
    Activations Density 0.038%

    No Known Activations