© 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-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 11180
    Prev
    Next
    INDEX
    Explanations

    March, climate, analogy, successful

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

    The neuron fires on formal “linking” words and adverbs (e.g. whether, however, various, via, liability, costs) that commonly appear as mid-clause connectors in legal or technical prose.

    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
     Acho
    0.77
    と思います
    0.73
    फारिश
    0.70
     pampered
    0.67
    と思いました
    0.67
    После
    0.66
    ে
    0.66
    置き
    0.66
    正確
    0.65
     Runde
    0.65
    POSITIVE LOGITS
    s
    0.94
    sites
    0.89
    split
    0.78
     તેમાં
    0.75
    sion
    0.75
    school
    0.74
    iveness
    0.71
    sing
    0.71
    substantial
    0.71
    sit
    0.71
    Activations Density 0.000%

    No Known Activations