© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 100775
    Prev
    Next
    INDEX
    Explanations

    following past tense verbs

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

    This neuron activates on words that start a new sentence or main clause (e.g. sentence-initial pronouns, articles, and existential “there”).

    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
     Hinsicht
    -1.45
     şeyler
    -1.39
     some
    -1.38
     asbestos
    -1.36
     appalling
    -1.35
     interestingly
    -1.33
    راف
    -1.32
     mainly
    -1.30
     duże
    -1.30
    </b>
    -1.29
    POSITIVE LOGITS
    了解更多
    1.49
    пасибо
    1.43
    桖
    1.28
     muñeco
    1.27
    にゃ
    1.27
     THEATRE
    1.26
    _____________
    1.23
     любую
    1.23
     لینک
    1.23
     fournisseur
    1.23
    Activations Density 0.032%

    No Known Activations