© 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. 15804
    Prev
    Next
    INDEX
    Explanations

    keywords ending a sequence

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

    The neuron fires most strongly on tokens that occur right at the end of each input segment.

    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
    Ғ
    1.01
    刈
    0.95
    т
    0.91
     Harvesting
    0.87
     introspection
    0.84
     dissection
    0.83
     Trav
    0.83
    ड़
    0.80
    কবি
    0.80
     cheg
    0.80
    POSITIVE LOGITS
    liert
    0.72
    ielten
    0.68
    duplicated
    0.68
    invalid
    0.67
    afe
    0.64
     失
    0.64
     zerstört
    0.64
     zerst
    0.63
     mornings
    0.62
    incorrect
    0.62
    Activations Density 0.000%

    No Known Activations