© 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. 34-GEMMASCOPE-RES-131K
    7. 116959
    Prev
    Next
    INDEX
    Explanations

    JAXB or Q Green Dic

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

    The neuron fires on blocks of uppercase technical identifiers and version strings (e.g. JAXB, XMLBinding, v1.0.6-b27).

    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
     Nail
    -0.77
    Nail
    -0.73
     Daf
    -0.73
     Kigali
    -0.72
     Carleton
    -0.70
     Melinda
    -0.69
    咖啡
    -0.69
    𝒽
    -0.68
     Sén
    -0.68
     Ivana
    -0.68
    POSITIVE LOGITS
    jax
    1.09
     jax
    1.00
    Jax
    0.97
     marshal
    0.93
     Jax
    0.93
     javax
    0.93
     JAXBElement
    0.91
     JAXB
    0.91
     Marsh
    0.87
     harika
    0.84
    Activations Density 0.022%

    No Known Activations