© 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. 2752
    Prev
    Next
    INDEX
    Explanations

    Date and Project

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

    The neuron primarily activates on numeric references and codes—years, dates, section or document numbers, and similar structured numerical identifiers.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    formal administrative or legal documents with reference numbers, dates, and institutional names.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    formal institutional or legal document references, especially structured metadata like case/report numbers, dates, acronyms, and official organization or conference titles.

    oai_token-act-pair · gpt-5Triggered 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
     kvůli
    0.91
    ពួកគេ
    0.82
     သူ့
    0.82
     తిన
    0.79
     দুর্দান্ত
    0.78
     दुनियाभर
    0.78
    पीड़न
    0.78
    ຫຼາຍ
    0.78
     મોટા
    0.77
     obwohl
    0.76
    POSITIVE LOGITS
     dated
    1.05
     undersigned
    0.93
     hereby
    0.89
     ……………………
    0.85
     …………
    0.85
     hereinafter
    0.84
     平成
    0.84
     duly
    0.84
    ‟
    0.83
     :
    0.82
    Activations Density 0.190%

    No Known Activations