© 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. 39570
    Prev
    Next
    INDEX
    Explanations

    an or a followed by a noun

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

    The neuron strongly activates on the standalone capital letter “A,” i.e. the indefinite article “A” used by itself.

    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
     other
    -1.50
     already
    -1.46
     not
    -1.33
     just
    -1.13
     lately
    -1.10
     drugi
    -1.09
     older
    -1.08
     hardly
    -1.04
     latter
    -1.02
     to
    -1.00
    POSITIVE LOGITS
    🤑
    1.35
    nemers
    1.31
    疍
    1.28
    ַּ
    1.27
    ❄️
    1.27
     disponibilités
    1.27
    şağı
    1.26
    וּ
    1.24
    ָּ
    1.23
     siitä
    1.23
    Activations Density 0.039%

    No Known Activations