© 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
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 1437
    Prev
    Next
    INDEX
    Explanations

    This neuron is a “content‐word” detector: it lights up strongly on semantically rich words (nouns, verbs, adjectives) rather than on function words.

    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
     क्षेत्रों
    0.75
     различных
    0.74
     これらの
    0.72
     others
    0.70
           
    0.64
     individuals
    0.64
                 
    0.64
     Các
    0.64
               
    0.63
     }[
    0.63
    POSITIVE LOGITS
     badass
    1.51
     paranoia
    1.48
     nerd
    1.41
     geek
    1.38
     nostalgia
    1.38
     heresy
    1.34
     nerdy
    1.34
     patriotism
    1.33
     sarcasm
    1.28
     overkill
    1.27
    Activations Density 1.196%

    No Known Activations