© 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. 99054
    Prev
    Next
    INDEX
    Explanations

    treatise, treated, treatable

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

    The neuron fires on verbs meaning “to treat/discuss/deal with” (e.g. trata, aborda, behandelt, treats, 讨) across different languages.

    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
    厹
    -2.52
    archiwizowane
    -2.38
    趼
    -2.28
    </h1>
    -2.16
    炱
    -2.14
    你来
    -2.13
    鮰
    -2.08
    ࿙
    -2.08
     vielen
    -2.02
     Sekret
    -2.00
    POSITIVE LOGITS
    “.
    2.25
    .”
    2.20
     is
    2.19
    
    2.14
    但他
    2.11
    be
    1.98
     ".";
    1.98
    但她
    1.91
     única
    1.90
    或
    1.89
    Activations Density 0.005%

    No Known Activations