© 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. 16168
    Prev
    Next
    INDEX
    Explanations

    If... then questions

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

    The neuron fires most strongly on unusually rare or out‐of‐distribution tokens—i.e. uncommon words or odd lexical forms.

    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
    ਐ
    0.84
    ข้อง
    0.84
    ในปี
    0.79
     บ้าง
    0.76
    메이션
    0.76
    ร้อง
    0.74
    瑶
    0.72
    VIS
    0.71
    IDField
    0.71
     peaked
    0.70
    POSITIVE LOGITS
    ی
    1.09
    tic
    0.95
    ти
    0.88
     womb
    0.84
    ture
    0.80
    tle
    0.80
    n
    0.78
    issants
    0.78
    ią
    0.77
    tól
    0.76
    Activations Density 0.001%

    No Known Activations