© 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. 24-GEMMASCOPE-2-RES-16K
    7. 12923
    Prev
    Next
    INDEX
    Explanations

    abilities and cognitive states

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

    The neuron detects words that name personal qualities, traits, or mental/physical states (e.g. ambition, confusion, superiority).

    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
    0.86
     !
    0.81
     .
    0.80
     simmering
    0.76
     strewn
    0.75
     innumerable
    0.74
     ,
    0.72
    1
    0.71
     ?
    0.71
     l
    0.68
    POSITIVE LOGITS
     abilities
    0.85
    ડિયા
    0.71
    ্যেষ্ঠ
    0.71
    医師
    0.70
     Fähigkeiten
    0.70
    㐄
    0.68
    受信
    0.68
    蘅
    0.68
    妿
    0.68
     કરવાની
    0.67
    Activations Density 0.002%

    No Known Activations