© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 129986
    Prev
    Next
    INDEX
    Explanations

    key-value pair separators

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

    The neuron is detecting the start of JSON object field definitions, i.e. the “{“ immediately followed by a quoted key and colon.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     is
    -1.68
     we
    -1.62
     directamente
    -1.52
     shouldn
    -1.50
     cependant
    -1.45
     conozco
    -1.40
     should
    -1.40
     SHOULD
    -1.39
     Should
    -1.34
     οποία
    -1.30
    POSITIVE LOGITS
    tifiable
    1.64
    itisation
    1.64
    ariats
    1.62
    áticos
    1.61
    hirts
    1.58
    cketed
    1.58
    ismo
    1.52
    ........
    1.52
    alties
    1.52
    ilty
    1.51
    Activations Density 0.174%

    No Known Activations