© 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. 13434
    Prev
    Next
    INDEX
    Explanations

    control measures or symptoms

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

    The neuron selectively activates on relatively uncommon, content-bearing words—especially longer or domain-specific terms—while ignoring generic short function 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
    Ironically
    0.64
     എന്നതാണ്
    0.62
     Ironically
    0.60
     نداشت
    0.60
     한다는
    0.58
     aslında
    0.58
     estaven
    0.56
    டப்பட்டது
    0.56
     maniac
    0.56
     Несмотря
    0.55
    POSITIVE LOGITS
     hãy
    2.80
     Hãy
    2.27
    Hãy
    2.14
     try
    1.84
    อย่า
    1.80
     อย่า
    1.77
     đừng
    1.76
    请
    1.73
     please
    1.73
    してください
    1.73
    Activations Density 1.126%

    No Known Activations