© 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. 22-GEMMASCOPE-RES-131K
    7. 98568
    Prev
    Next
    INDEX
    Explanations

    describing legal actions, convictions, or defendants

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

    The main thing this neuron does is identify legal case citation headings (court names, parties, reporters, volumes, and other citation 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_22/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
     will
    -1.36
    ști
    -1.27
     still
    -1.23
     such
    -1.22
     any
    -1.17
     to
    -1.12
     just
    -1.11
     this
    -1.11
    şti
    -1.11
    (){
    -1.10
    POSITIVE LOGITS
    有两个
    1.32
     cinto
    1.16
    ֠
    1.13
     весьма
    1.12
     alleged
    1.09
     kekinian
    1.09
    
    1.08
     encantador
    1.06
     conster
    1.06
     språk
    1.06
    Activations Density 0.659%

    No Known Activations