© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 27668
    Prev
    Next
    INDEX
    Explanations

    legislative acts and years

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

    The neuron strongly activates on four-digit numbers corresponding to years in a document.

    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
    あー
    -1.40
    為に
    -1.32
     então
    -1.29
    indicate
    -1.29
    ようになります
    -1.23
    際には
    -1.23
    ようになる
    -1.22
    ときは
    -1.21
     }}=\
    -1.20
    getYear
    -1.20
    POSITIVE LOGITS
     (
    1.50
     which
    1.34
    してました
    1.24
     whose
    1.24
     in
    1.17
    骸
    1.14
    しましたが
    1.13
    でしょうね
    1.13
    UCIÓN
    1.13
     نیروهای
    1.13
    Activations Density 0.022%

    No Known Activations