Hiroaki Hayashi
Research scientist at Salesforce Research. I build and study large language models: how to train them, how to make them generate code, and lately, how to figure out where they fall apart. Previously at Carnegie Mellon with Graham Neubig.
Interests
- Code-generating LLMs: training, inference, and the CodeGen family
- Where LLMs fail: multi-turn reasoning, long-context, evaluation gaps
- Agentic software engineering, LLMs as developers not just assistants
Selected Work
-
LLMs Get Lost In Multi-Turn Conversation
Philippe Laban*, Hiroaki Hayashi*, Yingbo Zhou, Jennifer Neville
-
CLOVER: A Test Case Generation Benchmark with Coverage, Long-Context, and Verification
Jiacheng Xu, Bo Pang, Jin Qu, Hiroaki Hayashi, Caiming Xiong, Yingbo Zhou
-
XGen-7B Technical Report
Erik Nijkamp*, Tian Xie*, Hiroaki Hayashi*, Bo Pang*, Congying Xia*, and others
-
CodeGen2.5: Small, but mighty (Blog)
Erik Nijkamp*, Hiroaki Hayashi*, Yingbo Zhou, Caiming Xiong
-
CodeGen2: Lessons for Training LLMs on Programming and Natural Languages
Erik Nijkamp*, Hiroaki Hayashi*, Caiming Xiong, Silvio Savarese, Yingbo Zhou
-
CodeGen: An Open Large Language Model for Code with Multi-Turn Program Synthesis
Erik Nijkamp*, Bo Pang*, Hiroaki Hayashi*, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong
-
Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processing
Pengfei Liu, Weizhe Yuan, Jinlan Fu, Zhengbao Jiang, Hiroaki Hayashi, Graham Neubig
-
WikiAsp: A Dataset for Multi-domain Aspect-based Summarization
Hiroaki Hayashi, Prashant Budania, Peng Wang, Chris Ackerson, Raj Neervannan, Graham Neubig
-
GSum: A General Framework for Guided Neural Abstractive Summarization
Zi-Yi Dou, Pengfei Liu, Hiroaki Hayashi, Zhengbao Jiang, Graham Neubig
Little Things