© 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. 337
    Prev
    Next
    INDEX
    Explanations

    multilingual conjunctions and examples

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

    This neuron fires on paragraph-opening or section-heading phrases—i.e. discourse markers that introduce a new point (e.g. “The role…,” “Special mention…,” “Speaking of…,” “As for…,” “Christmas…,” etc.).

    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
     Nhưng
    0.55
     Consequently
    0.54
     سپس
    0.50
     但是
    0.50
     此
    0.50
     بنابراین
    0.49
     그러나
    0.49
     Tetapi
    0.48
     但
    0.48
     Однако
    0.47
    POSITIVE LOGITS
     erstmal
    0.47
     вроде
    0.46
    꺄
    0.45
     yep
    0.44
    ğ
    0.44
     একদম
    0.44
    !),
    0.43
    పోయ
    0.43
     धमाकेदार
    0.43
    groomed
    0.43
    Activations Density 0.141%

    No Known Activations