© 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. 41400
    Prev
    Next
    INDEX
    Explanations

    json structure with keys

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

    The neuron is specialized in detecting numeric literals (numbers) in JSON or code-like text.

    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
     were
    -1.47
    </blockquote>
    -1.39
     doanh
    -1.34
    着他
    -1.33
     arī
    -1.32
    只是
    -1.30
    čno
    -1.29
    いなかった
    -1.29
     oferece
    -1.27
    飆
    -1.27
    POSITIVE LOGITS
     to
    1.30
     arbejds
    1.29
     Kasih
    1.25
    daging
    1.24
    税込
    1.23
    笑笑
    1.23
     spørgsmål
    1.23
     вчера
    1.22
     hjemmeside
    1.21
    ᕱ
    1.20
    Activations Density 1.491%

    No Known Activations