© 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. 98059
    Prev
    Next
    INDEX
    Explanations

    'd, l'apostrophe

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

    The neuron is highly tuned to pick up the French idiom “avoir l’air” (e.g. the tokens in “J’ai l’air …”), i.e. the “to look/seem” construction in French.

    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
     should
    -1.61
     which
    -1.60
     in
    -1.33
    !!!!!
    -1.32
     both
    -1.28
    !!!!
    -1.28
     on
    -1.23
     so
    -1.20
     just
    -1.18
     whose
    -1.18
    POSITIVE LOGITS
     ausein
    1.43
     aussit
    1.31
    ’
    1.30
    mathrm
    1.22
     dokter
    1.22
    푆
    1.21
     élastique
    1.21
    maske
    1.20
     apparaissent
    1.20
     exempl
    1.19
    Activations Density 0.013%

    No Known Activations