© 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. 7630
    Prev
    Next
    INDEX
    Explanations

    This neuron strongly responds to numbers, numeric measurements, units, and quantitative thresholds in the text.

    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
     twice
    0.57
     TW
    0.55
     trochu
    0.55
     Следу
    0.53
     Tang
    0.52
     Terrible
    0.52
     thrice
    0.52
     똑
    0.51
     잠깐
    0.51
     terrible
    0.50
    POSITIVE LOGITS
    超过
    2.45
    超過
    2.33
     exceeding
    2.22
    小于
    2.14
     exceed
    2.08
    大于
    2.00
    低于
    1.98
     exceeds
    1.96
     $>
    1.96
     >=
    1.89
    Activations Density 0.701%

    No Known Activations