© 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. 34-GEMMASCOPE-RES-131K
    7. 63191
    Prev
    Next
    INDEX
    Explanations

    judges' names before opinion

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

    The neuron flags tokens written largely or entirely in uppercase (e.g. court names, headings, and citation elements).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    compile
    -0.75
    Sew
    -0.71
    Heinz
    -0.69
    rdd
    -0.69
    पू
    -0.68
    queries
    -0.66
    比べると
    -0.65
     solutions
    -0.64
     udf
    -0.64
     Si
    -0.63
    POSITIVE LOGITS
     appellate
    1.97
     Appellate
    1.59
     appeals
    1.58
     Appeals
    1.56
     Appeal
    1.43
     appeal
    1.30
    Appe
    1.19
     unpublished
    1.16
     COA
    1.13
     APPELLATE
    1.11
    Activations Density 1.147%

    No Known Activations