© 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. 6072
    Prev
    Next
    INDEX
    Explanations

    origin and origins

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

    This neuron detects expressions of historical origin—phrases like “dating back to,” “traced back to,” or “going back to [time]” that indicate when something began.

    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
     dijadikan
    0.62
     แล
    0.60
     menghadapi
    0.58
     uzav
    0.57
     preso
    0.55
     आजीवन
    0.55
     সামনের
    0.54
     눌
    0.54
    ವಹ
    0.53
     หลัง
    0.52
    POSITIVE LOGITS
     origins
    2.56
    origins
    2.44
    起源
    2.38
    origin
    2.34
    Origins
    2.31
     origin
    2.28
     Origins
    2.22
    Origin
    2.17
     originates
    2.14
     originated
    2.09
    Activations Density 0.365%

    No Known Activations