© 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. 22-GEMMASCOPE-RES-131K
    7. 75133
    Prev
    Next
    INDEX
    Explanations

    Online Store

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

    The neuron responds to low‐level assembly loop code—especially NOP instructions and the small immediate numeric literals (e.g. “1”) used in delay loops.

    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
     sensi
    -1.12
     that
    -1.10
     Capítulo
    -1.09
     Índice
    -1.07
     sofá
    -1.06
    出来ない
    -1.04
     wios
    -1.02
     cordón
    -1.02
     created
    -1.01
     observer
    -1.01
    POSITIVE LOGITS
     $$$
    1.14
     nochmals
    1.13
    餉
    1.10
     AMAZING
    1.09
     fortifications
    1.08
    肆
    1.08
    战斗力
    1.07
     isolado
    1.06
     Carthag
    1.05
     evtl
    1.04
    Activations Density 0.000%

    No Known Activations