© 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. 23945
    Prev
    Next
    INDEX
    Explanations

    names and chemical compounds

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

    The neuron detects names with the element "Mohammad/Ahmad" and chemical compound structures containing "phenyl" groups.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron strongly activates on Arabic personal names—especially “Ahmad” (and to a lesser extent “Mohammad”)—in context.

    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
    ’
    -3.55
    蕻
    -2.81
    ſſen
    -2.77
    圌
    -2.63
    🧌
    -2.53
     celebrates
    -2.50
    憷
    -2.48
     knows
    -2.38
    霂
    -2.36
    籁
    -2.34
    POSITIVE LOGITS
    Surprisingly
    2.91
     Surprisingly
    2.78
    Interestingly
    2.77
     Interestingly
    2.64
    However
    2.64
    Because
    2.47
     Primarily
    2.47
    What
    2.47
     Ironically
    2.45
    Contrary
    2.45
    Activations Density 0.005%

    No Known Activations