© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 93037
    Prev
    Next
    INDEX
    Explanations

    system operation and functioning

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

    The neuron strongly activates on the word “operation,” i.e. it detects occurrences of “operation” in technical or procedural contexts.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     parere
    -1.65
     paziente
    -1.47
     commento
    -1.42
    ㉗
    -1.34
     récomp
    -1.32
     fisico
    -1.31
    菒
    -1.30
     disegno
    -1.30
    дання
    -1.29
    色んな
    -1.28
    POSITIVE LOGITS
     and
    1.97
     operation
    1.87
     opera
    1.58
     operates
    1.51
    絨
    1.50
     with
    1.49
     operational
    1.47
     data
    1.47
     operations
    1.46
     operating
    1.45
    Activations Density 0.060%

    No Known Activations