© 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. 44612
    Prev
    Next
    INDEX
    Explanations

    legal appeals

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

    This neuron consistently activates on different forms of the word “appeal,” detecting mentions of appeals or appealing.

    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
    .
    -1.80
     «
    -1.77
    In
    -1.64
     by
    -1.59
    </h2>
    -1.55
     ,
    -1.55
     (
    -1.55
    iyaki
    -1.53
    ">
    -1.44
     accuse
    -1.42
    POSITIVE LOGITS
    橚
    2.55
    赑
    2.44
    樒
    1.95
    栳
    1.93
    漚
    1.91
     短发
    1.89
     bendera
    1.88
     bedankt
    1.86
    枼
    1.86
     meravigli
    1.84
    Activations Density 0.064%

    No Known Activations