© 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. 25505
    Prev
    Next
    INDEX
    Explanations

    cross-appellant

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

    The neuron activates on court‐case party designations, i.e. labels like “Plaintiff,” “Defendant,” “Appellant,” and “Appellee.”

    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
     cose
    -1.05
    ڦ
    -1.04
     disser
    -0.97
     ordina
    -0.97
     compone
    -0.96
     coper
    -0.95
     cavo
    -0.93
     conector
    -0.93
    但却
    -0.90
     publica
    -0.89
    POSITIVE LOGITS
    istolet
    0.84
    ای
    0.80
    apit
    0.79
     komfort
    0.79
    arb
    0.78
    hydrat
    0.78
     diseñador
    0.78
     СССР
    0.78
    providedIn
    0.77
     linii
    0.77
    Activations Density 0.020%

    No Known Activations