© 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. 9688
    Prev
    Next
    INDEX
    Explanations

    encyclopedias and Britannica

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

    The neuron primarily detects numeric tokens (especially digits appearing in dates or years).

    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
     belanja
    0.56
    继
    0.54
    ござい
    0.54
     Yelp
    0.53
    我们需要
    0.53
    「
    0.53
     parro
    0.52
     „
    0.51
    仴
    0.51
     「
    0.50
    POSITIVE LOGITS
     Britannica
    1.15
    britann
    1.15
     Encyclopædia
    1.14
    Encycl
    0.98
     Encyclopaedia
    0.93
     britann
    0.82
    Brit
    0.79
     Encyclopedia
    0.72
     Britann
    0.71
     Facts
    0.70
    Activations Density 0.041%

    No Known Activations