© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 14380
    Prev
    Next
    INDEX
    Explanations

    What's up / What's wrong / What's it

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

    The neuron strongly activates on the first word of a spoken line—especially initial greetings or question words—marking the start of a character’s dialogue.

    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_24_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
    ↵
    0.84
    !
    0.78
     the
    0.65
    v
    0.64
    !!!
    0.62
     {
    0.61
    the
    0.60
     cultural
    0.60
      
    0.60
    !!
    0.60
    POSITIVE LOGITS
     районе
    0.84
    𝙤
    0.79
     pergunta
    0.78
     MDLVertex
    0.77
     BlueprintName
    0.75
     таком
    0.75
     النوم
    0.75
     Tentang
    0.73
     வகையான
    0.72
     чего
    0.72
    Activations Density 0.005%

    No Known Activations