© 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. 22-GEMMASCOPE-RES-131K
    7. 129356
    Prev
    Next
    INDEX
    Explanations

    This neuron detects programming calls that acquire or release resources—verbs like open, close, dispose, flush, restore, etc.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    closing or unlocking resources

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     inicialmente
    -1.06
     nuovo
    -1.00
    
    -0.98
     antigo
    -0.95
    avocat
    -0.94
     новой
    -0.92
    ptăm
    -0.92
     antiguo
    -0.91
     rebaj
    -0.91
     алгорит
    -0.91
    POSITIVE LOGITS
     after
    1.83
     or
    1.23
     Closing
    1.16
     outro
    1.05
     closing
    1.05
     when
    1.02
     afterwards
    1.02
     endIndex
    0.97
    AfterClass
    0.97
    أة
    0.96
    Activations Density 0.005%

    No Known Activations