© 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. 412
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates at the beginnings (“<bos>”) and ends (sentence-final punctuation) of sentences, effectively detecting sentence boundaries.

    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
    _
    0.29
    name
    0.28
    ().
    0.28
    _'
    0.28
     по
    0.28
     हुई
    0.28
    ['
    0.27
     задержа
    0.27
    ()?
    0.27
    ();
    0.26
    POSITIVE LOGITS
     Additionally
    0.54
     Furthermore
    0.53
     Moreover
    0.51
     Often
    0.49
     Inoltre
    0.49
     Ultimately
    0.47
     Also
    0.46
     Además
    0.45
     Even
    0.44
     Although
    0.44
    Activations Density 5.488%

    No Known Activations