© 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. 6195
    Prev
    Next
    INDEX
    Explanations

    Cambodian food, depression, quirky descriptions

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

    The neuron is a high‐frequency token detector—it lights up strongly on very common words (basic function words, pronouns, and other ultra-frequent tokens) rather than on specific content.

    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
     নেবার
    0.96
    实用
    0.94
     Dengan
    0.93
     thence
    0.93
     调用
    0.93
     cheerfully
    0.91
     despatch
    0.90
    i
    0.89
     この
    0.89
     দেবার
    0.88
    POSITIVE LOGITS
     badass
    1.16
     impactful
    1.13
     relatable
    1.11
     oftentimes
    1.10
     subpar
    1.05
     fanbase
    1.05
     metaverse
    1.04
     millennials
    1.03
     intertwined
    1.02
     ≠
    1.02
    Activations Density 0.519%

    No Known Activations