© 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. 31-GEMMASCOPE-2-RES-16K
    7. 15706
    Prev
    Next
    INDEX
    Explanations

    Eldoria within

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

    The neuron selectively detects mentions of characters and real‐world people—proper names and relational titles (like “Jim,” “Dr.,” “husband,” “mother,” 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_31_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
    俘
    0.68
     booty
    0.66
     thief
    0.63
     adventure
    0.62
     hostel
    0.60
     intruder
    0.60
     eğit
    0.60
     adventures
    0.59
     ujian
    0.59
    ंजा
    0.59
    POSITIVE LOGITS
     Madison
    0.86
    ██
    0.82
    Madison
    0.81
    Lincoln
    0.76
     Lincoln
    0.75
     fictional
    0.70
     Hawthorne
    0.70
     Stanton
    0.70
    American
    0.68
     Harper
    0.68
    Activations Density 0.184%

    No Known Activations