method · first seen 5 Oct 2026
group sequence policy optimization
Also: GSPO
Reinforcement learning algorithm optimizing sequence-level importance ratios for language model training.
Mentions per week
PapersCodeDiscussion
Earliest sources
- T-Search: An Open Agentic Retriever and Playground for Hard Multi-Step SearchPapers · 5 Oct 2026 · Olga Tsymboi, Ramil Latypov, Aleksandr Medvedev, Danil Taranets, Dmitrii Stoiano…
- Replan for Error Recovery: Training Embodied VLMs through Verifiable Interaction with Action-Grouped Turn-GSPOPapers · 19 Oct 2026 · Zhuoran Wang, Keqing Fan, Yang Li, Jinbo Su, Ke Wang, Sijia Luo, Yuanzheng Tan,…
Latest sources
- Replan for Error Recovery: Training Embodied VLMs through Verifiable Interaction with Action-Grouped Turn-GSPOPapers · 19 Oct 2026 · Zhuoran Wang, Keqing Fan, Yang Li, Jinbo Su, Ke Wang, Sijia Luo, Yuanzheng Tan,…
- T-Search: An Open Agentic Retriever and Playground for Hard Multi-Step SearchPapers · 5 Oct 2026 · Olga Tsymboi, Ramil Latypov, Aleksandr Medvedev, Danil Taranets, Dmitrii Stoiano…