태그: Google-Research

2026년 7월 21일

• Attention-free
• Efficient-Transformer
• FFT
• FNet
• Fourier-Transform
• Google-Research
• Linear-Complexity
• Token-Mixing

2026년 7월 21일

• Chain-of-Thought
• Flan-Collection
• Flan-T5
• Google-Research
• Instruction-Tuning
• MMLU
• Multitask-Fine-tuning
• T5
• Zero-Shot
• Zero-Shot-Generalization

2026년 7월 21일

• Expert-Capacity
• Google-Research
• load-balancing
• Mixture-of-Experts
• Sparse-MoE
• Switch-Routing
• Switch-Transformer
• Top-1-Routing

2026년 7월 21일

• C4-Dataset
• Google-Research
• Multitask-Learning
• Relative-Attention-Bias
• Span-Corruption
• T5
• Text-to-Text
• Unified-Framework

2026년 7월 21일

• Cross-lingual-Transfer
• Google-Research
• Large-Vocabulary
• mC4-Dataset
• mT5
• Multilingual
• Temperature-Sampling

2026년 7월 21일

• Encoder-Decoder
• Flan-UL2
• Google-Research
• Mixture-of-Denoisers
• Mode-Token
• R-Denoiser
• S-Denoiser
• UL2
• Unified-LM
• X-Denoiser