© 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. 40918
    Prev
    Next
    INDEX
    Explanations

    conveying information and messages

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

    The neuron detects occurrences of the verb “convey” (and its inflected forms like “conveys” or “conveying”), i.e. words meaning to express or transmit information.

    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
    ६
    -1.67
     dAtA
    -1.66
    ۴
    -1.58
     vart
    -1.52
    biendo
    -1.49
    ๖
    -1.49
     виправивши
    -1.48
    ви
    -1.45
    ри
    -1.43
     наших
    -1.41
    POSITIVE LOGITS
     through
    1.52
     what
    1.47
     all
    1.47
     convey
    1.45
     their
    1.45
     possible
    1.44
     different
    1.44
     selaku
    1.43
    他不
    1.41
     
    1.40
    Activations Density 0.174%

    No Known Activations