© 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. 12642
    Prev
    Next
    INDEX
    Explanations

    Laplacian regularization methods

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

    This neuron specializes in picking out technical academic jargon—particularly mathematical and dynamical‐systems terms commonly found in research paper titles.

    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
     password
    0.56
    opedia
    0.52
     algebraically
    0.51
     Password
    0.51
     confund
    0.49
     malaysia
    0.49
    Passwords
    0.49
    📵
    0.49
    password
    0.49
     dron
    0.49
    POSITIVE LOGITS
     Sobolev
    0.92
    lesssim
    0.88
     Spatial
    0.75
    )\|_{
    0.74
    Spatial
    0.74
    }\|_{
    0.72
     PDEs
    0.71
     پکڑ
    0.69
     spatial
    0.68
    spatial
    0.68
    Activations Density 0.054%

    No Known Activations