© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 6029
    Prev
    Next
    INDEX
    Explanations

    specific grammatical structures

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

    The neuron primarily picks up occurrences of the modal auxiliary “can” (and its negation “cannot”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     incontr
    0.85
     riun
    0.85
     comunità
    0.75
    desi
    0.75
    覦
    0.75
     anche
    0.74
     też
    0.73
     sodass
    0.73
     turb
    0.72
    っても
    0.72
    POSITIVE LOGITS
    brahim
    0.77
     withered
    0.77
    зне
    0.75
     heightened
    0.73
     respite
    0.71
    🩶
    0.71
     heighten
    0.70
    🌿
    0.70
    стые
    0.70
     overwhelm
    0.69
    Activations Density 0.000%

    No Known Activations