© 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. 31-GEMMASCOPE-2-RES-16K
    7. 565
    Prev
    Next
    INDEX
    Explanations

    start of list or array

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

    The neuron predominantly activates on numeric tokens and numeric literals (e.g. array indices, IDs, and other standalone numbers).

    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
    ेंद्रीय
    0.43
    ṓ
    0.41
    ôn
    0.40
    ôtels
    0.40
     जॉन
    0.39
    ösen
    0.39
     क्वाड्र
    0.39
    clarinet
    0.39
    ovir
    0.38
    ẖ
    0.38
    POSITIVE LOGITS
    _
    0.52
     !=
    0.49
    ->
    0.48
    の
    0.48
    _->
    0.47
    _{
    0.43
    의
    0.43
     ==
    0.43
    ['
    0.41
    =(
    0.41
    Activations Density 0.002%

    No Known Activations