© 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. 34-GEMMASCOPE-RES-131K
    7. 48499
    Prev
    Next
    INDEX
    Explanations

    dates and events

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

    The neuron selectively fires on numeric tokens—especially years, dates, and other digit sequences (e.g. “1,” “1952,” “1992”) and, to a lesser extent, section‐heading words—so its main role is spotting numbers in the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    covid
    -1.30
     covid
    -1.24
     Covid
    -1.13
    androidx
    -1.13
     microbiome
    -1.05
    Covid
    -1.03
    ​​​​
    -1.00
    ​​​​​​​
    -0.99
     COVID
    -0.93
    COVID
    -0.92
    POSITIVE LOGITS
    http
    1.07
     http
    1.06
     asfalto
    0.97
    Recently
    0.95
    カテゴリ
    0.87
     recently
    0.87
    ̼
    0.84
    模板
    0.82
    😏
    0.80
     Recently
    0.79
    Activations Density 0.233%

    No Known Activations