© 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. 63210
    Prev
    Next
    INDEX
    Explanations

    court case citations

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

    The neuron is detecting legal document identifiers, i.e. sequences of numbers and punctuation used in case or docket citations.

    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
     krim
    -1.27
     (
    -1.24
     these
    -1.14
     but
    -1.13
     before
    -1.13
    ならず
    -1.13
     aſ
    -1.10
     маса
    -1.09
     francisco
    -1.05
     yng
    -1.05
    POSITIVE LOGITS
     kultury
    1.25
     parku
    1.24
     samym
    1.19
     später
    1.16
     miasta
    1.12
     寶
    1.12
     odkry
    1.10
     zosta
    1.10
     Теперь
    1.09
     größer
    1.07
    Activations Density 0.106%

    No Known Activations