© 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. 32850
    Prev
    Next
    INDEX
    Explanations

    watching, seen, listen, buy

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

    The neuron detects first‐ and second‐person personal pronouns (e.g. I, me, my, you) indicating direct self‐reference or address.

    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
    reading
    -1.16
     reading
    -1.13
     Reads
    -1.13
     print
    -1.09
     Reading
    -1.06
     READ
    -1.04
     READING
    -1.03
     learn
    -1.00
     Learn
    -0.99
     readable
    -0.96
    POSITIVE LOGITS
     acronyms
    0.92
     thon
    0.91
    춘
    0.87
     вси
    0.82
     bawah
    0.77
    ключение
    0.76
     starte
    0.75
     Thon
    0.75
     स्टार
    0.75
     whofe
    0.75
    Activations Density 0.040%

    No Known Activations