© 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. 22-GEMMASCOPE-RES-131K
    7. 39907
    Prev
    Next
    INDEX
    Explanations

    the

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

    This neuron primarily detects the word “the,” especially when it appears in code comments or documentation.

    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
    Sabes
    -1.55
     TÉ
    -1.53
    躄
    -1.52
    könig
    -1.52
    Estás
    -1.51
     constat
    -1.48
    Nép
    -1.48
    Zubereitung
    -1.47
    𝐖
    -1.47
    vereignty
    -1.43
    POSITIVE LOGITS
     from
    1.91
     At
    1.53
     provides
    1.52
     one
    1.46
    '
    1.43
     with
    1.42
     as
    1.41
     Even
    1.36
     information
    1.34
     is
    1.33
    Activations Density 0.001%

    No Known Activations