© 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. 5334
    Prev
    Next
    INDEX
    Explanations

    lists item separators

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

    The neuron fires strongly on forum‐style discussion headings—that is, the quoted forum or board names (e.g. “Discussion in ‘MacBook Pro’”) that introduce thread titles.

    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
    만
    0.60
    ланган
    0.59
     imaginar
    0.57
    甥
    0.57
    વાસ
    0.56
    พื้น
    0.55
    랴
    0.55
    чно
    0.55
    купно
    0.54
    лянчук
    0.54
    POSITIVE LOGITS
    
    0.88
    、“
    0.77
    、『
    0.75
     Archivado
    0.66
     Lords
    0.64
     Derrick
    0.64
     Lord
    0.63
    هِ
    0.63
     δή
    0.63
    ”(
    0.62
    Activations Density 0.195%

    No Known Activations