© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 85384
    Prev
    Next
    INDEX
    Explanations

    que followed by a word

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

    The neuron strongly activates on the token “que,” i.e. the common Spanish/Portuguese conjunction/relative word.

    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
     They
    -3.36
    ،
    -3.03
    ↵↵
    -2.88
    。“
    -2.77
    .,
    -2.55
    屃
    -2.48
    ႏ
    -2.47
    Ᏺ
    -2.47
     незавер
    -2.44
     è
    -2.41
    POSITIVE LOGITS
    嘥
    2.50
    </tfoot>
    2.42
    も良い
    2.39
     خمس
    2.38
    氿
    2.31
    ܞ
    2.28
     pady
    2.23
    0
    2.23
     &
    2.22
     октя
    2.22
    Activations Density 0.003%

    No Known Activations