© 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. 2922
    Prev
    Next
    INDEX
    Explanations

    This neuron strongly activates on spans containing the second-person pronoun “you” followed by a form of “to be” (especially “are”) plus a negation (as in “you are not there”).

    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
    ようになる
    0.68
     becoming
    0.61
    되기
    0.61
    される
    0.57
    될
    0.57
     ලැබ
    0.55
    becomes
    0.55
     misch
    0.54
    igia
    0.54
     बनेंगे
    0.53
    POSITIVE LOGITS
     đang
    1.81
    正在
    1.80
     sedang
    1.68
    处于
    1.56
    อยู่ใน
    1.49
     находя
    1.42
     berada
    1.41
     перебу
    1.41
     находится
    1.37
     sitting
    1.35
    Activations Density 0.744%

    No Known Activations