© 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. 24-GEMMASCOPE-2-RES-16K
    7. 11661
    Prev
    Next
    INDEX
    Explanations

    The neuron is sensitive to words about noticing or perceiving—terms that signal observation or attention (e.g. notice, spot, see, watch, scrutinize).

    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
     costru
    0.88
     asuntos
    0.85
    💪
    0.85
     त्यासाठी
    0.84
     refundable
    0.84
    ייס
    0.84
     decisions
    0.83
     குறைக்க
    0.82
     strive
    0.82
     मंजूरी
    0.82
    POSITIVE LOGITS
    看到的
    1.57
     glimpses
    1.48
     thấy
    1.40
     увидеть
    1.36
    観察
    1.31
    观看
    1.30
     glimpse
    1.29
     nhìn
    1.28
     attentively
    1.28
    Watching
    1.27
    Activations Density 3.063%

    No Known Activations