© 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. 22-GEMMASCOPE-RES-131K
    4. 39784
    Prev
    Next
    INDEX
    Explanations

    parameters and arguments

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

    This neuron detects section or field labels in code/docs—words like “fields,” “Parameters,” “attributes,” or “Options” that name parameter/metadata sections.

    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
     is
    -1.10
     first
    -1.06
     if
    -1.06
    伉
    -1.02
     in
    -1.02
     mehrere
    -1.02
    各大
    -1.02
    k
    -0.98
     has
    -0.97
     result
    -0.94
    POSITIVE LOGITS
    さんから
    1.38
     parámetros
    1.27
     thoſe
    1.27
    Rispondi
    1.27
    álise
    1.18
     Paramètres
    1.15
    包括
    1.14
    בּ
    1.14
    lizes
    1.13
    malink
    1.12
    Activations Density 0.033%

    No Known Activations