© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 44790
    Prev
    Next
    INDEX
    Explanations

    states or manages systems

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

    The neuron strongly activates on the token “control” (i.e. it detects occurrences of the word “control”).

    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
    '
    -4.03
    [
    -2.92
    s
    -2.70
     стал
    -2.58
    ва
    -2.31
    (
    -2.22
    蹰
    -2.20
     
    -2.19
    section
    -2.16
     dann
    -2.08
    POSITIVE LOGITS
    趼
    3.16
    谞
    3.06
    媗
    3.00
    งิน
    2.94
    ”:
    2.91
    袮
    2.86
     всѣ
    2.84
    
    2.83
    蚴
    2.81
    
    2.81
    Activations Density 0.050%

    No Known Activations