© 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
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 99696
    Prev
    Next
    INDEX
    Explanations

    optimizing methods and performance

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

    The neuron fires on occurrences of “optim-” words—that is, any form of optimize/optimizing/optimization.

    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
     and
    -2.06
     in
    -2.00
    呶
    -1.93
    反抗
    -1.88
     hará
    -1.83
    
    -1.82
    大力
    -1.81
    跸
    -1.80
    がたくさん
    -1.80
    [\
    -1.77
    POSITIVE LOGITS
    橚
    2.34
    呖
    2.17
    ↵
    2.14
    峮
    2.13
    壠
    2.02
     маленьких
    2.00
     оптими
    1.97
    蛧
    1.93
     тщательно
    1.92
    芣
    1.92
    Activations Density 0.031%

    No Known Activations