HYDRA: Unifying Multi-modal Generation and Understanding via Representation-Harmonized Tokenization

Documenti analoghi

• MixGRPO: Unlocking Flow-based GRPO Efficiency with Mixed ODE-SDE
  di: Li, Junzhe, et al.
  Pubblicazione: (2025)
• Omni-View: Unlocking How Generation Facilitates Understanding in Unified 3D Model based on Multiview images
  di: Hu, JiaKui, et al.
  Pubblicazione: (2025)
• UniCom: Unified Multimodal Modeling via Compressed Continuous Semantic Representations
  di: Zhao, Yaqi, et al.
  Pubblicazione: (2026)
• Symbiotic-MoE: Unlocking the Synergy between Generation and Understanding
  di: Liu, Xiangyue, et al.
  Pubblicazione: (2026)
• High-Performance Temporal Reversible Spiking Neural Networks with $O(L)$ Training Memory and $O(1)$ Inference Cost
  di: Hu, JiaKui, et al.
  Pubblicazione: (2024)