© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 7698
    Prev
    Next
    INDEX
    Explanations

    proper nouns and specific descriptions

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

    The neuron strongly activates on structural metadata tokens—like years, numbered list markers, headings, and other document‐level numbers and symbols—i.e. it flags metadata or section‐header elements.

    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
     PFAS
    2.23
     impactful
    2.22
     microbiome
    2.22
     onboarding
    2.19
     TikTok
    2.16
     ২০২২
    2.14
     🥰
    2.13
     reimag
    2.06
     millennials
    1.97
     fintech
    1.96
    POSITIVE LOGITS
    
    1.66
     muß
    1.27
    http
    1.20
     চিকিত
    1.07
     изпол
    1.05
    􀂾
    1.05
    StringBuffer
    1.01
    －－
    0.97
    �
    0.93
    ﾕ
    0.91
    Activations Density 0.721%

    No Known Activations