© 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. 34-GEMMASCOPE-RES-131K
    7. 25894
    Prev
    Next
    INDEX
    Explanations

    possessive pronouns followed by nouns

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

    The neuron is highly sensitive to personal possessive pronouns (e.g. “my” and “his”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     egiten
    -1.12
     estado
    -0.92
     lugar
    -0.92
    Ж
    -0.91
    šče
    -0.90
    AppRoutingModule
    -0.90
     getItemId
    -0.89
    ́
    -0.88
     desde
    -0.88
    了一个
    -0.85
    POSITIVE LOGITS
     eyes
    1.45
     hands
    1.20
     feet
    1.16
     utmost
    1.11
     own
    1.11
     newfound
    1.09
    Then
    1.03
     sights
    1.02
     maten
    1.00
     emphasis
    0.98
    Activations Density 0.019%

    No Known Activations