© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 1857
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on specific named entities (proper nouns), such as brand names, company names, product names, or TV‐show titles.

    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
     його
    0.49
    ኪያ
    0.44
    มัน
    0.44
    それを
    0.43
    ICENSE
    0.43
     его
    0.43
    これを
    0.43
    🆚
    0.42
    γκεκρι
    0.42
     उसे
    0.41
    POSITIVE LOGITS
     etc
    0.88
    などの
    0.81
     and
    0.77
     등의
    0.75
    ，
    0.67
     등이
    0.66
    etc
    0.65
    などを
    0.65
     প্রভৃতির
    0.65
     등을
    0.64
    Activations Density 0.001%

    No Known Activations