© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 2404
    Prev
    Next
    INDEX
    Explanations

    sunshine

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

    The neuron fires on mentions of sun‐related, warm‐weather contexts—words like “sun,” “sunblock,” “swim,” “suit,” and similar heat/sun exposure terms.

    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
     Михай
    0.59
     tonight
    0.57
     peace
    0.55
     Peace
    0.55
     Christmas
    0.54
    פקי
    0.54
     педагоги
    0.54
    龚
    0.54
    ponin
    0.53
     violin
    0.53
    POSITIVE LOGITS
     sun
    2.64
     sunscreen
    2.59
     sunburn
    2.44
     güneş
    2.36
    太阳
    2.33
     солнце
    2.33
    太陽
    2.28
    sun
    2.27
    Sun
    2.25
     الشمس
    2.25
    Activations Density 0.624%

    No Known Activations