© 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. 11614
    Prev
    Next
    INDEX
    Explanations

    chocolate cake

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

    This neuron fires on step-by-step instructional language—especially imperatives and related words in recipe or “how-to” directions.

    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
     sauté
    0.71
     लह
    0.65
    煸
    0.64
     padassa
    0.63
     slaw
    0.63
     saute
    0.63
    菜
    0.62
    炒
    0.62
     Lenz
    0.61
     broth
    0.61
    POSITIVE LOGITS
     chocolate
    1.94
    Chocolate
    1.91
     Chocolate
    1.90
    🍫
    1.88
    chocolate
    1.86
     chocolates
    1.84
     चॉकलेट
    1.82
     chocol
    1.77
     chocolat
    1.71
    チョコレート
    1.70
    Activations Density 0.889%

    No Known Activations