PlanSightRAG is a visual-first multimodal retrieval‑augmented generation system designed to automate question answering and compliance checking of civil standard plans. It processes plan imagery directly, using a ColNomic‑3B multi‑vector retrieval engine, an agentic Planner‑Retriever‑Auditor‑Synthesizer, and MaxSim heatmaps to provide an evidence trail. The system achieves high recall on a new 4,056‑pair benchmark from five state DOTs, and demonstrates near‑perfect verdict accuracy on synthetic compliance drawings when a rule threshold is supplied, outperforming OCR‑based baselines.
arXiv:2610.08102v1 Announce Type: new Abstract: Conversational MLLM agents are increasingly expected to assist in professional workflows, from AI research and engineering design to product management...
arXiv:2610.33277v1 Announce Type: new Abstract: Engineering standards are challenging knowledge sources for large language models (LLMs) because they are hierarchical, multimodal, and densely connect...
The paper introduces TerraAgent, a training‑free framework that enhances remote sensing visual question answering by integrating three modules: GDR for demonstration retrieval, MACE for evidence‑typed sub‑question decomposition, and CVA for counterfactual evidence auditing. It demonstrates significant improvements on multiple benchmarks, raising accuracy, visual‑dependency scores, and reducing prior‑only answers and hallucinations without fine‑tuning.
arXiv:2512. 11995v2 Announce Type: replace-cross Abstract: While many vision-language models (VLMs) are developed to answer well-defined, straightforward questions with highly specified targets, as in most benchmarks, they often struggle in practice with complex open-ended tasks, which usually require multiple rounds of exploration and reasoning in the visual space.
WeAgent-MMSearch introduces a multimodal search agent that preserves retrieved images as persistent references, enabling the model to inspect, process, and cite them throughout a search trajectory. The system includes a harness (WeAgent-Harness), a post‑training method (FA‑GSPO) that recovers salvageable rollouts, and a new benchmark (VisTarget‑Bench) to evaluate image‑retrieval versus visual‑perception failures. Evaluation shows that agentic post‑training boosts performance by 19.22 points, allowing the model to outperform similarly sized open‑source models and compete with much larger ones.
ATP‑Bench proposes a new benchmark for evaluating agentic tool planning in multimodal large language models (MLLMs) that generate interleaved text-and-image responses. The benchmark contains 7,702 QA pairs, including 1,592 visual‑question‑answer pairs, across eight categories and 25 visual‑critical intents, all verified by humans. A Multi‑Agent MLLM‑as‑a‑Judge (MAM) system is introduced to assess tool‑call precision, missed opportunities, and overall response quality without relying on ground‑truth references.
arXiv:2609.37349v1 Announce Type: cross Abstract: Multi-step visual retrieval-augmented generation (RAG) answers complex questions by repeatedly retrieving visual evidence, updating an intermediate s...
arXiv:2604.07146v3 Announce Type: replace Abstract: Knowledge-based visual question answering (KB-VQA) requires vision-language models to understand images and use external knowledge, especially for...
The paper introduces the Necessary Tool‑Evidence Path (NTEP) annotation scheme and its associated reward mechanism (NTEP‑R) to better supervise vision‑language models that use external tools. By explicitly specifying which evidence is needed and penalizing redundant tool calls, the authors train an 8B‑parameter model that shows improved accuracy and tool‑use efficiency across seven image‑grounded benchmarks. The approach demonstrates that fine‑grained supervision of tool‑evidence paths is essential for robust agentic VLM performance.
arXiv:2607. 22643v1 Announce Type: new Abstract: Multimodal retrieval-augmented generation (mRAG) aims to answer image-text queries with external knowledge, but most existing systems still retrieve directly from raw multimodal input over a flat evidence space.
StateSight is a new benchmark designed to isolate and evaluate the ability of vision‑language models to reconstruct latent spatial structure from a single image. It consists of three procedurally generated task families—cube‑net opposite‑face reasoning, occluded cube‑tower counting, and 4‑neighbor connected‑component counting—each with 300 deterministic prompts and exact‑match scoring. The benchmark also includes a companion dataset, StateSight‑Steps, with 900 image‑text examples and 3,600 intermediate visual states to aid analysis of reconstruction errors.
The paper extends a retrieval‑augmented generation (RAG) method to recover traceability links between industrial requirements that include both text and images. By retrieving text and image candidates separately, classifying each pair with a vision‑language model, and filtering links by verbalized confidence, the approach improves F1 scores on industrial datasets (0.310 vs. 0.114 for a baseline) while maintaining performance on public text‑only datasets. The study highlights that cross‑modal retrieval remains a bottleneck and that confidence filtering can trade precision for recall.