© 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. 687
    Prev
    Next
    INDEX
    Explanations

    diverse languages and categories

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

    The neuron strongly activates on individual terms when they appear as items in comma-separated enumerations (especially inside parentheses).

    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
     này
    0.36
    (
    0.35
     этой
    0.32
    л
    0.32
    àm
    0.31
     thuộc
    0.31
     two
    0.31
     neither
    0.30
     cervix
    0.30
     mà
    0.30
    POSITIVE LOGITS
    甚至是
    0.50
    その他の
    0.43
    なども
    0.43
     Bahkan
    0.42
    อื่นๆ
    0.40
     zelfs
    0.40
    甚至
    0.39
    その他
    0.39
     whatnot
    0.38
    <unused2130>
    0.37
    Activations Density 0.926%

    No Known Activations