© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 53955
    Prev
    Next
    INDEX
    Explanations

    prosecutor, state, appellant, police, people

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

    This neuron activates on tokens that are part of legal‐case metadata—like uppercase party names, court names, citation numbers, and other header information in legal documents.

    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
     berbagai
    -1.24
     mencionado
    -1.02
     nacido
    -0.98
     various
    -0.98
    相性
    -0.95
     panggilan
    -0.94
     setelah
    -0.94
     makan
    -0.93
    Alemania
    -0.93
    (\%)
    -0.92
    POSITIVE LOGITS
     Ofrece
    1.05
     ninh
    1.05
     tý
    1.02
    wanted
    0.96
     социали
    0.95
     کال
    0.92
     Lleva
    0.92
     хочет
    0.92
     continuare
    0.92
     эмоциона
    0.92
    Activations Density 0.201%

    No Known Activations