© 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. 55953
    Prev
    Next
    INDEX
    Explanations

    SWT dialogs and eclipse editors

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

    The neuron is tuned to pick out programming‐style identifiers—especially Java/SWT class and package names (e.g. “SWT,” “Dialog,” or dotted package paths).

    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
     quête
    -0.72
     Tsche
    -0.69
     hum
    -0.69
     mechanical
    -0.68
     humming
    -0.67
     collector
    -0.67
    hdc
    -0.66
     aff
    -0.66
     dull
    -0.65
    amela
    -0.65
    POSITIVE LOGITS
    Shell
    1.13
    Composite
    1.05
     Shell
    1.05
     Composites
    1.02
    MessageBox
    0.99
    shell
    0.89
     Composite
    0.85
    Sw
    0.85
    composite
    0.85
    SHELL
    0.82
    Activations Density 0.025%

    No Known Activations