© 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. 60269
    Prev
    Next
    INDEX
    Explanations

    learning and activities

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

    The neuron detects verbs that describe planned or scheduled activities, especially those following “will” (e.g. will play, will examine, will discuss, will learn).

    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
     this
    -1.34
     you
    -1.24
     that
    -1.20
    пите
    -1.17
    only
    -1.16
     who
    -1.16
    com
    -1.15
     encuentras
    -1.13
    deli
    -1.11
    per
    -1.09
    POSITIVE LOGITS
    minente
    1.23
    σιμοποι
    1.17
     avancé
    1.16
     gemeinsamen
    1.15
    EMIA
    1.13
     スクリーン
    1.11
    icidio
    1.11
    ttet
    1.10
    ANCEL
    1.09
     conjunta
    1.09
    Activations Density 0.056%

    No Known Activations