© 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. 45219
    Prev
    Next
    INDEX
    Explanations

    names after titles

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

    The neuron strongly activates on the formal header information of legal cases—especially proper names, titles, courts, and “of”/“Court of” phrases that together form the case caption metadata.

    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
    graffiti
    -0.77
     urdu
    -0.74
    昭和
    -0.74
     highschool
    -0.73
     iran
    -0.72
    ेम
    -0.71
     ary
    -0.69
     japanese
    -0.69
    鄄
    -0.69
     adsor
    -0.68
    POSITIVE LOGITS
     ceilings
    0.75
    thèse
    0.75
    泣き
    0.75
     fills
    0.74
     fill
    0.72
     شرح
    0.72
     accue
    0.71
    小麦
    0.71
    قاط
    0.71
    China
    0.69
    Activations Density 0.104%

    No Known Activations