MACER decouples cross-modal token-budget allocation from within-modality token ranking and improves accuracy over shared top-K compression in Qwen2.5-Omni and OmniVinci models.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Allocation Before Ranking: Decoupled Token Compression for OmniLLMs
MACER decouples cross-modal token-budget allocation from within-modality token ranking and improves accuracy over shared top-K compression in Qwen2.5-Omni and OmniVinci models.