© 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. 39221
    Prev
    Next
    INDEX
    Explanations

    which one

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

    The neuron selectively fires on the word “one,” especially when it appears in question phrases (e.g. “which one…,” “one…”) indicating a choice.

    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
     and
    -2.00
    こんばん
    -1.69
    ;
    
    -1.66
    谆
    -1.62
    ,”
    -1.57
    蚍
    -1.54
    )=\
    -1.53
    ”]
    -1.52
    我們
    -1.50
     servidores
    -1.49
    POSITIVE LOGITS
     he
    1.80
     she
    1.62
    した場合
    1.52
    骤
    1.51
    的一天
    1.49
    盏
    1.46
    いいのか
    1.44
     today
    1.43
     namun
    1.41
     мәкал
    1.40
    Activations Density 0.009%

    No Known Activations