© 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. 34-GEMMASCOPE-RES-131K
    7. 84237
    Prev
    Next
    INDEX
    Explanations

    own, Eigen

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

    The neuron is strongly selective for the orthographic strings “Eigen” (as in the Eigen linear‐algebra library) and “own” (as in ownership/get‐own‐property), firing whenever a token contains either of those substrings.

    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
    ^{*}(
    -0.90
     שני
    -0.84
    ürn
    -0.84
    Bridget
    -0.80
    ubakar
    -0.79
    <tfoot>
    -0.79
    erevan
    -0.79
    тому
    -0.79
    miguel
    -0.79
     cantidad
    -0.79
    POSITIVE LOGITS
     belongings
    1.01
     hábit
    1.00
     initiatives
    0.94
     initiative
    0.90
    ógico
    0.89
     crafted
    0.89
    资
    0.88
     давно
    0.85
    vedo
    0.85
    犠
    0.85
    Activations Density 0.012%

    No Known Activations