© 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. 11259
    Prev
    Next
    INDEX
    Explanations

    access and retrieval

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

    The neuron strongly activates on numeric tokens in parenthetical or bibliographic contexts (e.g. years, volume and page numbers), effectively detecting citation‐style references.

    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
    回避
    0.60
    慢慢
    0.57
    しながら
    0.54
    满
    0.54
     ali
    0.54
    နဲ့
    0.53
    過ぎ
    0.53
     ful
    0.52
     doom
    0.51
    ファイ
    0.51
    POSITIVE LOGITS
     Retrieved
    1.43
     Accessed
    1.42
     accessed
    1.27
     retrieved
    1.22
    Accessed
    1.18
    accessed
    1.09
     Diakses
    1.08
     สืบค้นเมื่อ
    1.03
     Consultado
    0.97
     Retrieve
    0.96
    Activations Density 0.193%

    No Known Activations