© 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. 2813
    Prev
    Next
    INDEX
    Explanations

    cannot and will not fulfill

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

    The neuron fires on emphatic legal quantifier terms—especially “any,” “all,” or the phrase “any and all” in formal regulatory or license text.

    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
     ferrovi
    0.44
    更是
    0.44
     "-",
    0.44
     rosem
    0.42
     addirittura
    0.42
     riff
    0.42
     flannel
    0.41
     pensando
    0.41
    POE
    0.41
    -
    0.41
    POSITIVE LOGITS
     مختلفة
    0.57
     unexplained
    0.56
    新た
    0.55
    長い
    0.55
     واحدة
    0.54
     unspecified
    0.53
     unexplored
    0.52
     unjustified
    0.51
     unexpected
    0.50
     unsuccessful
    0.50
    Activations Density 0.145%

    No Known Activations