5 papers
SoftmaxGRPO: Learning to Reason using Softmax Advantage Group Estimation
Jefferson Hernandez, Jaywon Koo, Zilin Xiao +2
Group-based reinforcement learning objectives such as GRPO can allocate learning signal poorly across prompt difficulty: under binary rewards, group normalization induces a diverge…
MetaEmbed: Scaling Multimodal Retrieval at Test-Time with Flexible Late Interaction
Zilin Xiao, Qi Ma, Mengting Gu +4
Universal multimodal embedding models have achieved great success in capturing semantic relevance between queries and candidates. However, current methods either condense queries a…
ProxyThinker: Test-Time Guidance through Small Visual Reasoners
Zilin Xiao, Jaywon Koo, Siru Ouyang +3
Recent advancements in reinforcement learning with verifiable rewards have pushed the boundaries of the visual reasoning capabilities in large vision-language models (LVLMs). Howev…
RAST: Reasoning Activation in LLMs via Small-model Transfer
Siru Ouyang, Xinyu Zhu, Zilin Xiao +3
Reinforcement learning (RL) has become a powerful approach for improving the reasoning capabilities of large language models (LLMs), as evidenced by recent successes such as OpenAI…
LOCORE: Image Re-ranking with Long-Context Sequence Modeling
Zilin Xiao, Pavel Suma, Ayush Sachdeva +4
We introduce LOCORE, Long-Context Re-ranker, a model that takes as input local descriptors corresponding to an image query and a list of gallery images and outputs similarity score…