© 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. 10-GEMMASCOPE-RES-131K
    7. 19014
    Prev
    Next
    INDEX
    Explanations

    originating source

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

    The neuron detects mentions of an entity’s provenance—tokens like “originating,” “originated,” or “origin” that signal where something comes from.

    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
    sistence
    -1.66
    hibited
    -1.51
    镲
    -1.48
    vertebrates
    -1.41
    formatie
    -1.41
    cursions
    -1.39
    cisive
    -1.39
    compati
    -1.39
    <bos>
    -1.38
     baumwolle
    -1.36
    POSITIVE LOGITS
     dotyczące
    1.59
     These
    1.53
     There
    1.47
     Because
    1.45
     meningkat
    1.41
     као
    1.41
    増加
    1.39
    鴛
    1.38
     lantaran
    1.38
     From
    1.38
    Activations Density 0.036%

    No Known Activations