© 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. 5886
    Prev
    Next
    INDEX
    Explanations

    rule explanation

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

    The neuron is specialized to pick up on strong negative judgment words (e.g. “unacceptable”).

    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
     увели
    0.79
     odborn
    0.74
    が増
    0.72
     offerings
    0.71
    ‌پ
    0.71
     предлагает
    0.71
     nabí
    0.71
     определяет
    0.71
     például
    0.71
     принима
    0.69
    POSITIVE LOGITS
     nurse
    0.81
    Painting
    0.78
    来
    0.78
    塞尔
    0.76
     questi
    0.75
     sutures
    0.73
     commemorative
    0.73
    게
    0.72
    可以
    0.72
    erian
    0.71
    Activations Density 0.000%

    No Known Activations