© 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. 36211
    Prev
    Next
    INDEX
    Explanations

    United States or U.S. officer

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

    The neuron strongly activates on the all‐caps headings and formal terms in appellate court documents (e.g. U.S., COURT, FILED, docket numbers, etc.).

    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
    verlag
    -1.00
    iffer
    -0.86
    meisterschaften
    -0.82
    pid
    -0.81
     amateurs
    -0.81
     Affili
    -0.78
    unikasi
    -0.77
    chymal
    -0.77
    ronomy
    -0.77
    dic
    -0.76
    POSITIVE LOGITS
     wpływ
    0.85
    acité
    0.85
     amikor
    0.85
     Überblick
    0.82
     sabido
    0.82
    ÁL
    0.81
     Konzentration
    0.80
     poème
    0.79
    妈
    0.77
    "},
    0.77
    Activations Density 0.027%

    No Known Activations