© 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. 64812
    Prev
    Next
    INDEX
    Explanations

    to + verb

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

    The neuron almost exclusively fires on the word “to” (especially when it’s introducing an infinitive verb phrase).

    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
    damento
    -1.13
     إليها
    -1.09
     абсолю
    -1.06
     RÉ
    -1.05
     так
    -1.05
    铠
    -1.02
    kowej
    -1.01
    ämään
    -1.00
    1
    -0.99
     more
    -0.99
    POSITIVE LOGITS
     warrant
    1.58
     τὸν
    1.36
     τὴν
    1.33
     hopes
    1.27
     μὲν
    1.23
    かもしれませんが
    1.20
     некоторы
    1.18
     appear
    1.18
     require
    1.17
     τῆς
    1.15
    Activations Density 0.024%

    No Known Activations