© 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. 9379
    Prev
    Next
    INDEX
    Explanations

    names or subjects followed by descriptions

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

    The neuron flags rare or domain-specific content words—uncommon nouns, proper names, and technical terms that carry topical information.

    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
    Preparación
    -0.97
     bénéficient
    -0.96
     Той
    -0.94
    sprzedam
    -0.92
     frappé
    -0.90
     aventures
    -0.89
     parlement
    -0.87
     broderie
    -0.87
    Pacote
    -0.86
    Dirección
    -0.85
    POSITIVE LOGITS
     her
    1.08
     she
    1.07
     intervenciones
    0.84
     級
    0.83
     allowing
    0.83
     với
    0.78
    孩子的
    0.75
     creencias
    0.75
     others
    0.75
     helping
    0.75
    Activations Density 0.074%

    No Known Activations