© 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. 444
    Prev
    Next
    INDEX
    Explanations

    as a/an/the

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

    The neuron is essentially detecting the standalone token “as.”

    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
     maniere
    -1.20
    讓人
    -1.16
    ビックリ
    -1.14
    让人
    -1.13
    浓郁
    -1.09
     všetky
    -1.04
    舒适
    -1.02
    often
    -1.02
     cardin
    -1.02
     canne
    -1.01
    POSITIVE LOGITS
     a
    2.13
     an
    1.65
     what
    1.57
     as
    1.55
     its
    1.38
     something
    1.37
     their
    1.37
     one
    1.31
     being
    1.25
     having
    1.17
    Activations Density 0.023%

    No Known Activations