© 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. 12918
    Prev
    Next
    INDEX
    Explanations

    Pathfinder, facebook research, LPIC

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

    This neuron fires strongly on words that occur at the very start of a sentence (i.e. the first token after a sentence boundary).

    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
    ྪ
    0.56
     באמצעות
    0.52
    Lieutenant
    0.52
    MongoClient
    0.50
    Governor
    0.50
    ólito
    0.50
    ᚗ
    0.50
    columnIndex
    0.49
    ContextMenu
    0.49
    pSensor
    0.49
    POSITIVE LOGITS
    /
    0.70
     crappy
    0.66
     specs
    0.60
     only
    0.59
     good
    0.59
     very
    0.58
     shipped
    0.57
     ebay
    0.57
     really
    0.57
     +/-
    0.56
    Activations Density 0.003%

    No Known Activations