© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 34030
    Prev
    Next
    INDEX
    Explanations

    multilingual conjunctions and qualifiers

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

    The neuron activates on multi-character CJK (Chinese/Japanese) word fragments, i.e. non-Latin script tokens.

    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
    獁
    -1.51
    –
    -1.47
    Angebot
    -1.43
    foundland
    -1.42
    cartão
    -1.40
     ↓,
    -1.39
     Inoltre
    -1.36
    mice
    -1.33
    titial
    -1.33
     发型
    -1.33
    POSITIVE LOGITS
     albeit
    1.51
    ではある
    1.45
     returns
    1.38
     situations
    1.37
    且
    1.33
    ’
    1.31
     enough
    1.31
    とはいえ
    1.29
     jednak
    1.27
     looking
    1.27
    Activations Density 0.002%

    No Known Activations