© 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. 75652
    Prev
    Next
    INDEX
    Explanations

    Ahmad and Ahmed names

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

    The neuron strongly activates on proper personal names—especially Arabic names like Ahmad/Ahmed (and variants such as Ahmet or Mahmoud).

    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
     "
    -3.20
    凊
    -2.16
     bahwa
    -2.14
     idyllic
    -2.14
     ("
    -2.14
     These
    -2.13
     dichas
    -2.09
    I
    -2.09
     siete
    -2.05
    。
    -2.03
    POSITIVE LOGITS
    \
    2.92
    ፑ
    2.50
    》。
    2.36
    ).”
    2.34
     …”
    2.28
    徧
    2.25
    鄣
    2.22
    その他の
    2.22
    揷
    2.16
    塒
    2.13
    Activations Density 0.003%

    No Known Activations