© 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. 68545
    Prev
    Next
    INDEX
    Explanations

    PRIVATE KEY

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

    The neuron strongly activates on specialized technical identifiers and jargon—e.g. function or type names, module/component labels, and domain-specific acronyms in code.

    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
     composto
    -1.05
    Ekster
    -1.03
     indeb
    -0.98
    var
    -0.90
     oscu
    -0.88
     proveniente
    -0.88
     aquesta
    -0.88
     suscit
    -0.86
    这么
    -0.86
     sobri
    -0.86
    POSITIVE LOGITS
    znej
    1.07
     decidido
    0.96
     ONLY
    0.95
     inspección
    0.94
    leistung
    0.94
     tonel
    0.94
     или
    0.93
     Inclusion
    0.93
     illustratie
    0.91
     laiton
    0.91
    Activations Density 0.072%

    No Known Activations