© 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-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 59867
    Prev
    Next
    INDEX
    Explanations

    German words and names

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

    The neuron responds to occurrences of the substring “dank” (in any casing) as a standalone token.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    ↵↵
    -3.25
    </i>
    -2.64
    %;
    
    -2.61
    もちろん
    -2.36
     тщательно
    -2.27
     awaken
    -2.27
     skillfully
    -2.23
     atractiva
    -2.23
     greats
    -2.19
     внимательно
    -2.19
    POSITIVE LOGITS
    0
    3.95
     berdua
    2.77
    各样的
    2.67
    t
    2.61
    是这样
    2.47
    o
    2.31
    埽
    2.31
     -----
    2.31
    cited
    2.27
    7
    2.23
    Activations Density 0.003%

    No Known Activations