© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 5032
    Prev
    Next
    INDEX
    Explanations

    German numbers and articles

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

    This neuron activates on action-oriented instruction words or directives (e.g., Please, expand, recommend, testing, built), marking steps in procedural or technical instructions.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
    ко
    0.70
    ↵↵
    0.69
     ,
    0.69
     ridiculously
    0.69
    <start_of_image>
    0.67
    煕
    0.66
    出来る
    0.66
     joka
    0.66
    тари
    0.66
    кар
    0.65
    POSITIVE LOGITS
     ampere
    1.65
     adenine
    1.63
     einige
    1.16
     einigen
    1.16
     jeder
    1.09
     magie
    1.05
     einer
    1.04
     bilden
    1.04
    䞍
    1.02
     nope
    1.01
    Activations Density 0.031%

    No Known Activations