© 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. 131027
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on tokens that signal law‐practice or attorney names and titles (e.g. “Offices,” “P.C.,” “lawyer,” “firm,” “Attorney,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    Legal professionals and firms

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     regiões
    -1.32
     Cantidad
    -1.28
     escolas
    -1.28
     koszt
    -1.27
     attraverso
    -1.25
     propriedades
    -1.25
    ά
    -1.21
     configurações
    -1.20
     camadas
    -1.18
     mesmas
    -1.16
    POSITIVE LOGITS
     lawyer
    1.47
     attorney
    1.45
    savevideo
    1.35
    ִּ
    1.34
     law
    1.34
    büro
    1.28
     médec
    1.27
    шком
    1.26
     경우
    1.25
     cases
    1.24
    Activations Density 0.037%

    No Known Activations