© 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. 1442
    Prev
    Next
    INDEX
    Explanations

    modelproduct model names

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

    This neuron detects specific bicycle model names and their numeric identifiers (e.g. “Giant Defy 5,” “Diamondback Haanjo Tero,” “Airen 1,” etc.).

    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
     ascertaining
    0.65
    foaf
    0.60
    <unused491>
    0.60
     fruitful
    0.59
     fraught
    0.57
    tAux
    0.57
    糺
    0.57
    ज्ञापन
    0.57
    INTa
    0.57
    慌
    0.57
    POSITIVE LOGITS
     Series
    1.21
     Pro
    1.12
     series
    1.08
    シリーズ
    0.98
    series
    0.97
     Elite
    0.97
     Plus
    0.96
    系列的
    0.95
     II
    0.94
    Series
    0.93
    Activations Density 0.615%

    No Known Activations