© 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. 31-GEMMASCOPE-2-RES-16K
    4. 15911
    Prev
    Next
    INDEX
    Explanations

    calling when cold

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

    This neuron strongly fires on the word “call” (in its various forms: call, called, “call to”) – it’s essentially detecting that token.

    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
    imenta
    0.57
    CurrentImage
    0.56
     Stability
    0.54
    
    0.54
    েনার
    0.52
    recognized
    0.52
     regiones
    0.52
     recognized
    0.52
     wsz
    0.52
    יום
    0.52
    POSITIVE LOGITS
     冷
    0.59
     конкурен
    0.58
     फ्लिप
    0.55
    agi
    0.54
    überkü
    0.54
     냉
    0.52
     Delaware
    0.51
     delusions
    0.51
    atlar
    0.51
    abelle
    0.50
    Activations Density 0.001%

    No Known Activations