© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 13878
    Prev
    Next
    INDEX
    Explanations

    technical documentation and articles

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

    The neuron activates on occurrences of “shared” (e.g. in shared_ptr or shared module/import paths), i.e. it’s looking for references to shared code or shared data structures.

    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
    ↵
    -2.47
    .
    -2.20
     neuen
    -1.93
     With
    -1.87
     juga
    -1.84
     Also
    -1.84
    明年
    -1.82
     bekämp
    -1.71
     ก็
    -1.70
     The
    -1.69
    POSITIVE LOGITS
     uly
    2.05
    笮
    1.99
    𞤢
    1.96
     istan
    1.95
     bors
    1.95
     dezelve
    1.94
     kwanza
    1.84
     soos
    1.84
     hals
    1.84
     mily
    1.84
    Activations Density 0.027%

    No Known Activations