An agentic mRAG framework uses GRPO-trained visual reranking and active rejection to verify retrieved candidate entities, achieving state-of-the-art on three KB-VQA benchmarks.
beyond token-level answer equivalence for question answering evaluation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
MMAgent-R$^2$: Learning to Rerank and Reject for Agentic mRAG
An agentic mRAG framework uses GRPO-trained visual reranking and active rejection to verify retrieved candidate entities, achieving state-of-the-art on three KB-VQA benchmarks.