© 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. 64805
    Prev
    Next
    INDEX
    Explanations

    possessive pronouns followed by nouns

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

    This neuron detects idiomatic prepositional phrases expressing viewpoint (e.g. “to my surprise,” “for reasons known only to her”).

    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
     kyt
    -1.27
     chaleco
    -1.26
     炖
    -1.26
    tifies
    -1.20
    vereignty
    -1.18
    foundland
    -1.17
    óricas
    -1.17
    haviours
    -1.16
     compagnon
    -1.16
    ásticas
    -1.15
    POSITIVE LOGITS
    ,
    1.56
    They
    1.23
    '
    1.18
    Depending
    1.17
    Despite
    1.15
    "
    1.15
    Once
    1.12
    /***
    1.11
    While
    1.10
    One
    1.09
    Activations Density 0.042%

    No Known Activations