© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 75867
    Prev
    Next
    INDEX
    Explanations

    she when what where whom

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

    The neuron is strongly triggered by numeric or quantitative expressions—years, ages, dates, numbers, and other figure‐based tokens.

    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
     esso
    -2.77
    He
    -2.45
    They
    -2.08
     îl
    -1.91
    彼らの
    -1.88
     dessen
    -1.87
     usarlo
    -1.68
    彼の
    -1.63
    他們的
    -1.62
     ihnen
    -1.61
    POSITIVE LOGITS
     she
    8.88
     her
    5.69
     она
    4.91
    她
    4.69
     вона
    3.75
     herself
    3.64
     ее
    3.52
     She
    3.50
    เธอ
    3.50
     היא
    3.33
    Activations Density 0.407%

    No Known Activations