© 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. 12543
    Prev
    Next
    INDEX
    Explanations

    optional punctuation

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

    This neuron spikes on uncommon, information‐dense tokens (i.e. rare full‐word pieces) rather than on common subword fragments.

    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
     කැ
    0.86
    hug
    0.82
    翟
    0.80
     виконав
    0.79
    سی
    0.78
    说
    0.78
    ді
    0.78
    哲学
    0.77
     פּ
    0.77
    javax
    0.75
    POSITIVE LOGITS
     favorably
    0.70
     favorable
    0.69
    mäß
    0.66
    alities
    0.64
     assuring
    0.63
     предыду
    0.62
    ার
    0.62
    iliency
    0.62
    лые
    0.61
     précédentes
    0.61
    Activations Density 0.000%

    No Known Activations