© 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. 11178
    Prev
    Next
    INDEX
    Explanations

    code delimiters and logic relations

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

    The neuron fires on little “headline” phrases that name something (often a product or room) followed immediately by an evaluative descriptor (e.g. “Room was comfortable,” “Slip covers for dining room chairs…,” “Windmill aerator kit up…,” “The furniture and other online offers…”).

    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
     stencil
    0.68
     neuer
    0.67
     rusted
    0.65
     epoxy
    0.65
     cruc
    0.64
     euthan
    0.63
     готовы
    0.63
     chitosan
    0.61
     propane
    0.61
     aquare
    0.59
    POSITIVE LOGITS
     Πολ
    0.70
     במהלך
    0.69
     Система
    0.64
     Findings
    0.63
    Ти
    0.61
    Когда
    0.61
     Tijdens
    0.61
     Podczas
    0.61
     During
    0.61
     Automated
    0.61
    Activations Density 0.010%

    No Known Activations