today local_bar

Quyen Tran

AI Resident, Qualcomm Inc, QualComm

1 paper at NeurIPS 2025

Homepage· OpenReview· Semantic Scholar· Google Scholar

Poster Session 6

Friday, December 5, 2025 · 4:30 PM → 7:30 PM

Exhibit Hall C,D,E

Token-Level Self-Play with Importance-Aware Guidance for Large Language Models

#3309 · Tue Le, Hoang Tran Vuong, Quyen Tran, Linh Ngo Van, Mehrtash Harandi, Trung Le

We introduce SWIFT, a token-level self-alignment and distillation method that improves LLMs by assigning token weights via a teacher model