© 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. 12550
    Prev
    Next
    INDEX
    Explanations

    Here's why

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

    The neuron strongly activates on standalone punctuation tokens (commas, dashes, etc.) rather than on words.

    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
     sincerely
    0.87
    Personally
    0.82
    在我
    0.81
    Honestly
    0.81
     Honestly
    0.80
     Personally
    0.77
     Sincerely
    0.75
     我
    0.73
     Frankly
    0.72
     Sebagai
    0.71
    POSITIVE LOGITS
     unless
    1.01
    除非
    1.00
    unless
    0.96
     look
    0.93
     compare
    0.82
    look
    0.81
     try
    0.81
     notice
    0.77
     ancak
    0.74
     Unless
    0.74
    Activations Density 0.110%

    No Known Activations