© 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
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 10290
    Prev
    Next
    INDEX
    Explanations

    repeating process or task

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

    The neuron fires on past-tense (–ed) verbs that report discrete events or actions, as typically found in news‐style descriptions.

    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
    öp
    0.48
    RCP
    0.48
    odo
    0.48
     కొన
    0.48
    asma
    0.47
    пол
    0.46
     günü
    0.46
    awi
    0.46
    ट्टी
    0.45
    рей
    0.45
    POSITIVE LOGITS
     twice
    4.00
    twice
    3.66
     Twice
    3.52
     thrice
    3.36
    多次
    3.30
    两次
    3.27
    几次
    3.25
     zweimal
    2.94
    何度も
    2.92
     repeatedly
    2.88
    Activations Density 0.650%

    No Known Activations