© 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-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 1169
    Prev
    Next
    INDEX
    Explanations

    drinks consumed with activities

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

    The neuron fires strongly on words naming beverages (e.g., coffee, cocoa, Coke, lemonade).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Gr
    -1.01
    gr
    -0.99
    喝酒
    -0.89
    就是这样
    -0.88
     drinking
    -0.87
    AZ
    -0.86
     arbeiten
    -0.85
     concernant
    -0.84
    al
    -0.84
    g
    -0.82
    POSITIVE LOGITS
     while
    1.31
     before
    1.08
     whilst
    0.88
     gemüt
    0.88
     to
    0.88
     purchased
    0.85
     and
    0.84
     served
    0.84
    &=&\
    0.84
    rasında
    0.83
    Activations Density 0.015%

    No Known Activations