© 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. 34-GEMMASCOPE-RES-131K
    4. 82823
    Prev
    Next
    INDEX
    Explanations

    Xinjiang, Uighur, Chengdu, Lake

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

    The neuron is highly selective for tokens containing numeric sequences (e.g. years, numeric IDs, digits).

    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
     насељу
    -0.80
     Lebanese
    -0.74
     createDate
    -0.71
    えない
    -0.71
    檬
    -0.69
     Swanson
    -0.68
    氡
    -0.68
    青年
    -0.68
    卵
    -0.68
    索引
    -0.67
    POSITIVE LOGITS
     Xinjiang
    1.79
     Uy
    1.59
    Xin
    1.20
    Uy
    1.13
    ウイ
    1.00
     Kash
    0.94
     Josh
    0.87
     Liu
    0.87
     Ü
    0.86
     Xin
    0.84
    Activations Density 0.007%

    No Known Activations