© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 15176
    Prev
    Next
    INDEX
    Explanations

    making computers do things

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

    This neuron fires on broad “aside” or meta‐comment clauses—sentences that give a high‐level caveat, variation, or instruction (e.g. “Interesting variation on the trolley problem,” “the most popular way … is to determine whether …,” “Other accessories not included,” “Made in USA,” etc.).

    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
    attività
    0.78
    endereco
    0.78
    
    0.77
     літера
    0.76
    ۶
    0.75
    Table
    0.75
    алет
    0.75
    に係る
    0.74
     هڪ
    0.74
    CommandBuffer
    0.73
    POSITIVE LOGITS
     wavy
    0.89
     villages
    0.80
     styled
    0.77
     bumpy
    0.75
     goal
    0.74
     styles
    0.73
     checkpoints
    0.73
     bars
    0.72
     gyms
    0.72
     picket
    0.72
    Activations Density 0.001%

    No Known Activations