© 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. 9114
    Prev
    Next
    INDEX
    Explanations

    text followed by sx

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

    This neuron primarily fires on uppercase alphanumeric tokens and technical acronyms—including model numbers and other identifiers combining letters, digits, and punctuation.

    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
    물
    0.59
    ත්ව
    0.59
    okan
    0.58
    ਰ
    0.56
    haus
    0.54
    的情況
    0.54
    oes
    0.54
    picture
    0.53
    nection
    0.52
    nut
    0.51
    POSITIVE LOGITS
     finalText
    0.60
     cetera
    0.59
    ябре
    0.57
    がる
    0.56
     тексто
    0.56
     বিষ
    0.54
     ET
    0.54
    হাসিক
    0.54
     Tutt
    0.53
     etiqueta
    0.53
    Activations Density 0.690%

    No Known Activations