© 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. 9576
    Prev
    Next
    INDEX
    Explanations

    court case numbers

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

    The neuron fires on numeric tokens that are layout or page‐coordinate markers (e.g. line numbers or vertical position values) in the document formatting.

    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
    geç
    -1.13
    årige
    -0.93
    きたいと思います
    -0.93
    Válasz
    -0.93
     recinto
    -0.90
    了一会儿
    -0.89
    Élet
    -0.87
    Ū
    -0.85
    gambarkan
    -0.84
     They
    -0.84
    POSITIVE LOGITS
    🟪
    1.10
    vraagd
    1.02
    entieth
    0.97
     informaron
    0.95
    完成です
    0.93
     ciertas
    0.91
    马车
    0.91
     delantero
    0.91
     quedará
    0.91
     dijeron
    0.91
    Activations Density 0.001%

    No Known Activations