© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 11879
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on the very first words of new sentences or paragraphs—i.e. on sentence/segment-initial tokens.

    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_31_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
     YOUR
    0.84
     YOU
    0.84
    !"
    0.74
     Hopefully
    0.71
    !</
    0.71
     सफलतापूर्वक
    0.71
    ！
    0.68
    !")
    0.68
     "!
    0.67
     ！
    0.66
    POSITIVE LOGITS
     juxtaposition
    1.00
     reminded
    0.98
     analogy
    0.84
     reread
    0.83
     juxt
    0.82
    偶然
    0.82
     revisiting
    0.82
     pondering
    0.82
     juxtap
    0.81
     analogies
    0.80
    Activations Density 0.308%

    No Known Activations