© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 61351
    Prev
    Next
    INDEX
    Explanations

    to do

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

    This neuron primarily detects the word “do” (and its occurrence in the infinitive “to do”).

    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
    了很多
    -1.73
     mahdol
    -1.70
    粜
    -1.69
     folgende
    -1.64
    鱻
    -1.63
    ーん
    -1.60
    ウナ
    -1.59
    льзя
    -1.59
    我就是
    -1.59
    ーーー
    -1.55
    POSITIVE LOGITS
     with
    2.02
     in
    1.80
     for
    1.57
     from
    1.56
     They
    1.55
     Because
    1.55
    一流
    1.52
     and
    1.51
    歳の
    1.48
     when
    1.48
    Activations Density 0.012%

    No Known Activations