© 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. 6041
    Prev
    Next
    INDEX
    Explanations

    condition or purpose followed by a noun

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

    The neuron strongly activates on numeric or quantitative tokens—digits, numbered references (e.g. “Art. 6 para. 1”), and other explicit number-or code-like mentions.

    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
     -
    0.53
    p
    0.52
    7
    0.50
    1
    0.49
    0
    0.47
    #
    0.45
     ơn
    0.45
    b
    0.44
     №
    0.44
    m
    0.43
    POSITIVE LOGITS
     Hermitian
    0.49
     protagonists
    0.42
     অ্যাপ্লিকেশন
    0.42
     HStack
    0.41
     Gestaltung
    0.41
     কীভাবে
    0.40
     básicamente
    0.39
    それを
    0.39
    レシ
    0.39
     salespeople
    0.38
    Activations Density 0.217%

    No Known Activations