Hyung Won Chung
Research Scientist, OpenAI | Co-Creator of Flan-T5 & Instruction-Tuning Pioneer
Research Scientist at OpenAI and formerly lead research scientist at Google Brain. Lead author of the milestone Flan-T5 and Flan-PaLM papers, which proved that instruction fine-tuning across thousands of diverse NLP tasks dramatically improves model zero-shot reasoning and generalization. Deep expert in large-scale distributed pretraining, inference dynamics, and foundation model alignment.
Areas of focus
Professional niches
Proof of Work
Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
Co-authored the foundational paper discovering that generating step-by-step reasoning tokens before the final answer dramatically unlocks multi-step arithmetic, commonsense, and symbolic reasoning in foundation models.
Chain-of-thought benefits appear as an emergent property only in models with sufficient scale (~100B+ parameters or instruction-tuned compact models).
Context: PaLM 540B, GPT-3 175B tested on GSM8K math benchmark and SVAMP.
View missionScaling Instruction-Finetuned Language Models (Flan-T5 & Flan-PaLM)
Spearheaded the research demonstrating that instruction fine-tuning scaled across 1,836 diverse tasks and 540 billion parameters enhances task-general reasoning, chain-of-thought problem solving, and zero-shot performance across every benchmark.
Instruction tuning without careful task balance can degrade pure memorization capacity on open-domain fact retrieval.
Context: Flan-T5 and Flan-PaLM models evaluated on MMLU, Big-Bench, and HumanEval.
View mission