© 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. 28634
    Prev
    Next
    INDEX
    Explanations

    next or other

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

    The neuron is triggered by common English function words (especially “the” and “of” often seen in titles or lists) and associated punctuation.

    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
    behind
    -0.97
    (/[
    -0.96
     Ser
    -0.95
     เพ
    -0.89
     предоставля
    -0.88
     BEHIND
    -0.85
    doc
    -0.83
    었습니다
    -0.82
    着自己的
    -0.81
    졌
    -0.81
    POSITIVE LOGITS
     other
    1.88
     another
    1.50
     below
    1.42
     next
    1.41
    อื่น
    1.41
     later
    1.30
     hereinafter
    1.23
    次の
    1.20
     بعدی
    1.18
     suivante
    1.16
    Activations Density 0.022%

    No Known Activations