© 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. 12110
    Prev
    Next
    INDEX
    Explanations

    code blocks and specific symbols

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

    The neuron is keyed to section‐heading or document‐title tokens—i.e. the words that kick off paper titles, datasheet headings or other major document sections.

    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
     *)&
    0.75
    FormProvider
    0.74
     אס
    0.73
     혹은
    0.71
    ᄊ
    0.68
    theless
    0.67
    tains
    0.66
    Invalid
    0.66
    ())));
    0.66
    стю
    0.65
    POSITIVE LOGITS
    utsu
    0.87
     deuteron
    0.85
    唓
    0.82
     Dexter
    0.82
    ik
    0.80
     hedon
    0.78
    یک
    0.77
    aham
    0.76
    anaconda
    0.76
    ី
    0.76
    Activations Density 0.000%

    No Known Activations