© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 26165
    Prev
    Next
    INDEX
    Explanations

    personal pronouns and possessives

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

    The neuron is strongly activated by common function words—especially high‐frequency prepositions, pronouns, and conjunctions (e.g. “to,” “by,” “but,” “has,” “I,” “would”).

    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
     cosy
    -0.85
     viendo
    -0.84
     menyiapkan
    -0.83
    visející
    -0.79
    UG
    -0.79
    chargez
    -0.79
    हे
    -0.78
     hiszen
    -0.78
    Fig
    -0.78
     nouveautés
    -0.77
    POSITIVE LOGITS
     me
    2.27
     ours
    2.05
     mine
    1.82
     us
    1.76
     my
    1.53
     nosotros
    1.51
     мной
    1.47
     conmigo
    1.41
     мне
    1.34
     benim
    1.33
    Activations Density 0.130%

    No Known Activations