© 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. 13260
    Prev
    Next
    INDEX
    Explanations

    terms of endearment

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

    This neuron detects direct audience addresses and salutations (e.g. “Dear …”).

    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
    hra
    0.61
    ́c
    0.61
    PLEX
    0.60
     সর্বনিম্ন
    0.60
     ACL
    0.59
    __."
    0.59
     descom
    0.57
    ί
    0.57
    ंपल
    0.57
    fair
    0.56
    POSITIVE LOGITS
     precious
    1.00
     Precious
    0.87
    貴重
    0.86
     dearly
    0.84
     dear
    0.82
    dear
    0.79
     treasured
    0.72
     dearest
    0.71
     дорого
    0.71
     Dear
    0.71
    Activations Density 0.151%

    No Known Activations