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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 12485
    Prev
    Next
    INDEX
    Explanations

    2020

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

    The neuron fires on multi-digit year‐like number tokens (e.g. “2020,” “200200,” etc.). It’s essentially detecting year-style numeric strings.

    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
     Dokter
    -2.14
    𞤢
    -2.03
     sustent
    -2.03
    哢
    -2.03
    锨
    -2.02
     VARIOUS
    -2.00
     Energía
    -1.97
     dezelve
    -1.91
    憮
    -1.90
    臵
    -1.90
    POSITIVE LOGITS
    You
    3.30
    y
    2.58
     of
    2.58
    p
    2.45
    or
    2.36
    v
    2.31
    k
    2.30
    a
    2.28
    In
    2.23
    There
    2.23
    Activations Density 0.007%

    No Known Activations