© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 3527
    Prev
    Next
    INDEX
    Explanations

    This neuron responds strongly to the opening content words of new sections or headings—i.e. the first capitalized word(s) that introduce a paragraph, title, or subheading.

    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
     என்னுடைய
    0.95
    非常的
    0.91
    Notwithstanding
    0.89
    或者是
    0.83
    につきまして
    0.82
     непосред
    0.80
     Notwithstanding
    0.79
     весьма
    0.79
    또한
    0.75
     великолеп
    0.75
    POSITIVE LOGITS
     montrent
    0.72
     vznik
    0.67
     různ
    0.64
     praz
    0.62
     الخميس
    0.62
     rougeâtres
    0.61
     katva
    0.61
     koristi
    0.61
     dhamme
    0.61
     jaundice
    0.60
    Activations Density 0.028%

    No Known Activations