© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 65316
    Prev
    Next
    INDEX
    Explanations

    States court

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

    The neuron strongly activates on uppercase headings and names in legal documents—e.g. court titles, attorney labels, and party names in the case caption.

    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
    をご覧ください
    -0.98
    どうやら
    -0.92
    readString
    -0.90
    Added
    -0.87
    closer
    -0.86
    を読
    -0.85
     οπο
    -0.84
    好不容易
    -0.84
    ction
    -0.83
    追記
    -0.83
    POSITIVE LOGITS
    家に
    0.93
    Puck
    0.90
     кухне
    0.88
     mengunjungi
    0.85
     Азии
    0.83
     at
    0.82
     chords
    0.81
     menghubungi
    0.80
     fredag
    0.79
    𝑹
    0.78
    Activations Density 0.005%

    No Known Activations