© 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. 34-GEMMASCOPE-RES-131K
    7. 59895
    Prev
    Next
    INDEX
    Explanations

    metadata keys: value

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

    The neuron fires on words used as keys in structured data or front‐matter (i.e. tokens immediately preceding a colon in YAML/JSON/markup configurations).

    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
    你这个
    -1.02
    为啥
    -0.94
    你这
    -0.94
    事前に
    -0.87
     スペシャル
    -0.85
    ą
    -0.85
     ロゴ
    -0.84
     họ
    -0.84
    你来
    -0.84
     Poincar
    -0.83
    POSITIVE LOGITS
     “
    1.30
     both
    1.20
     zowel
    1.09
    !***
    1.07
    ABAD
    1.07
    DeclareMath
    1.06
    ```
    1.05
    detek
    1.02
    terapia
    0.99
     zarówno
    0.99
    Activations Density 0.133%

    No Known Activations