© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 13498
    Prev
    Next
    INDEX
    Explanations

    737 MAX jet

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

    The neuron detects standalone numeric tokens—especially three‐digit model or product numbers.

    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
    =
    -1.99
     ideally
    -1.95
    '
    -1.95
     For
    -1.93
     I
    -1.92
    的我
    -1.89
    7
    -1.82
    AfterViewInit
    -1.81
    6
    -1.81
    0
    -1.80
    POSITIVE LOGITS
    蜴
    2.20
    珰
    2.17
    nitř
    2.14
    adakan
    2.11
    蔸
    2.06
    ŚĆ
    2.03
    髕
    1.99
    짭
    1.96
     炖
    1.92
    瑷
    1.92
    Activations Density 0.067%

    No Known Activations