© 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. 14268
    Prev
    Next
    INDEX
    Explanations

    1920s, 1930s, 1940s, 1950s, 1960s

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

    This neuron activates on the numerals marking 20th-century years (e.g., the “19__s” decades like 1920s, 1930s).

    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
     mignon
    -1.72
     eau
    -1.62
    derdag
    -1.59
     visse
    -1.59
     ſy
    -1.57
    就连
    -1.55
     pérd
    -1.52
    istically
    -1.52
     catast
    -1.51
    嘮
    -1.49
    POSITIVE LOGITS
     gminy
    1.78
     just
    1.74
    okko
    1.60
     powi
    1.60
    (!__
    1.59
     但
    1.55
     level
    1.55
     curious
    1.51
     near
    1.49
     eager
    1.49
    Activations Density 0.018%

    No Known Activations