← Back to People
Researcher working on alignment, reasoning, and post-training for large language models at IBM Research
1
Mentions
1
Articles
1
Outlets
Articles
Most recent firstMentioned In
Referenced in coverage
New training method cuts LLM reasoning tokens by 11x using abstract token chains
arXivpositiveApr 30, 2026
“Lead author of Abstract Chain-of-Thought paper showing how to reduce LLM reasoning tokens by 11.6x using abstract token chains.”