CiteVQA requires models to cite specific document regions with bounding boxes alongside answers and finds that even the strongest MLLMs frequently cite the wrong region, with top SAA scores of only 76.0 for closed models and 22.5 for open-source ones.
hub
Visrag: Vision-based retrieval-augmented generation on multi-modality documents
17 Pith papers cite this work. Polarity classification is still indexing.
17
Pith papers citing it
hub tools
citation-role summary
background 1
citation-polarity summary
years
2026 17roles
background 1polarities
background 1representative citing papers
DocPrune is a training-free token pruning method that removes background and irrelevant tokens from document images using question and comprehension signals, yielding 3x encoder and 3.3x decoder throughput gains plus +1 F1 on M3DocRAG.
Bottleneck Tokens paired with a masked generative objective achieve state-of-the-art unified multimodal retrieval performance among 2B-scale models on the MMEB-V2 benchmark with 78 datasets.
ColChunk adaptively chunks visual document patches into contextual multi-vectors via clustering, cutting storage by over 90% while raising average nDCG@5 by 9 points.
VISOR is a unified agentic VRAG framework with Evidence Space structuring, visual action evaluation/correction, and dynamic sliding-window trajectories trained via GRPO-based RL that achieves SOTA performance on long-horizon visual reasoning benchmarks.
MARVEL reaches 37.9 nDCG@10 on the MM-BRIGHT benchmark by combining LLM query expansion, a reasoning-enhanced dense retriever, and GPT-4o CoT reranking, beating prior multimodal encoders by 10.3 points.
PLUME uses latent-state autoregressive rollouts and a progressive training curriculum to deliver efficient reasoning for universal multimodal embeddings without generating explicit rationales.
VLADriver-RAG reaches a new state-of-the-art Driving Score of 89.12 on Bench2Drive by retrieving structure-aware historical knowledge through spatiotemporal semantic graphs and Graph-DTW alignment.
Rewrite-driven generation with alignment and RL produces shorter, more effective generative multimodal embeddings than CoT methods on retrieval benchmarks.
POINTS-Seeker-8B is an 8B multimodal model trained from scratch for agentic search that uses seeding and visual-space history folding to outperform prior models on six visual reasoning benchmarks.
SLQ adapts frozen MLLMs for multimodal retrieval by appending shared latent queries to text and image tokens and introduces KARR-Bench to test knowledge-aware reasoning retrieval.
HIVE raises multimodal retrieval nDCG@10 to 41.7 on the MM-BRIGHT benchmark by inserting LLM-driven hypothesis generation and verification between retrieval passes, delivering +9.5 over the best text-only baseline and +14.1 over the best multimodal baseline.
FileGram grounds AI agent personalization in file-system behavioral traces via a data simulation engine, a diagnostic benchmark, and a bottom-up memory architecture.
MicroWorld constructs a multimodal attributed property graph from scientific image-caption data and augments MLLM prompts via retrieval to raise Qwen3-VL-8B performance by 37.5% on MicroVQA and 6% on MicroBench.
DocSeeker improves long-document understanding in MLLMs via a two-stage training process that combines supervised fine-tuning from distilled data with evidence-aware group relative policy optimization and memory-efficient resolution allocation.
BRIDGE reaches 29.7 nDCG@10 on MM-BRIGHT by RL-aligning multimodal queries to text and using a reasoning retriever, beating multimodal encoders and, when combined with Nomic-Vision, exceeding the best text-only retriever at 33.3.
Advanced language representations shape LLMs' schemas to improve knowledge activation and problem-solving.
citing papers explorer
-
MARVEL: Multimodal Adaptive Reasoning-intensiVe Expand-rerank and retrievaL
MARVEL reaches 37.9 nDCG@10 on the MM-BRIGHT benchmark by combining LLM query expansion, a reasoning-enhanced dense retriever, and GPT-4o CoT reranking, beating prior multimodal encoders by 10.3 points.
-
HIVE: Query, Hypothesize, Verify An LLM Framework for Multimodal Reasoning-Intensive Retrieval
HIVE raises multimodal retrieval nDCG@10 to 41.7 on the MM-BRIGHT benchmark by inserting LLM-driven hypothesis generation and verification between retrieval passes, delivering +9.5 over the best text-only baseline and +14.1 over the best multimodal baseline.
-
BRIDGE: Multimodal-to-Text Retrieval via Reinforcement-Learned Query Alignment
BRIDGE reaches 29.7 nDCG@10 on MM-BRIGHT by RL-aligning multimodal queries to text and using a reasoning retriever, beating multimodal encoders and, when combined with Nomic-Vision, exceeding the best text-only retriever at 33.3.