© 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. 34-GEMMASCOPE-RES-131K
    4. 55727
    Prev
    Next
    INDEX
    Explanations

    algebra and arithmetic concepts

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

    The neuron is highly sensitive to the word "algebra" and its morphological variants (e.g. "algebraic," "algebraically").

    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
     emphatically
    -0.88
     desaparición
    -0.79
    柔ら
    -0.79
     wisely
    -0.78
     ignited
    -0.78
     apri
    -0.77
     expressly
    -0.77
     correndo
    -0.75
     richly
    -0.75
     عزی
    -0.74
    POSITIVE LOGITS
     manipulative
    0.94
     operations
    0.91
    *);
    0.90
     MÉXICO
    0.89
     manipulations
    0.87
    jg
    0.86
    aI
    0.86
    ائرة
    0.86
    zda
    0.85
    Qb
    0.85
    Activations Density 0.033%

    No Known Activations