© 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
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 5742
    Prev
    Next
    INDEX
    Explanations

    Miyoko's Kitchen

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

    This neuron primarily signals paragraph or section boundaries (i.e. the blank‐line or break between text blocks).

    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
    ény
    0.61
     एंप
    0.57
    스타
    0.54
     Cuando
    0.53
     Informal
    0.52
     đoạn
    0.52
     INDIANA
    0.52
    ᒍ
    0.52
     Dipl
    0.52
     Nebraska
    0.52
    POSITIVE LOGITS
     socio
    0.57
     parameter
    0.55
    孅
    0.54
     ডাঃ
    0.53
    亦
    0.51
     resonance
    0.51
    CharPtr
    0.51
     soc
    0.50
     пациен
    0.50
     multifunctional
    0.49
    Activations Density 0.035%

    No Known Activations