Retrieval-augmented generation (RAG) systems must balance retrieval granularity with contextual coherence, a challenge that existing methods address through LLM-guided chunking, single-level context expansion, or hierarchical summarization. These approaches variously depend on costly LLM calls during indexing or retrieval, limit context aggregation to a single granularity level, or introduce information loss through summarization. We present SproutRAG, an attention-guided hierarchical RAG framework that builds a binary tree over sentence-level chunks using learned inter-sentence attention from a sentence-level large language model (SLLM). To avoid proximity bias from uniform attention averaging, SproutRAG learns head- and layer-level aggregation weights that identify attention patterns most aligned with semantic document structure. At retrieval time, it uses hierarchical beam search over the tree to gather candidates at multiple granularities, capturing multi-sentence relevance beyond flat retrieval. The framework is trained end-to-end with a joint objective that improves both embeddings and tree structure, without requiring external LLM calls. Experiments across four benchmarks spanning scientific, legal, and open-domain settings demonstrate that SproutRAG improves information efficiency (IE) by 6.1% on average over the strongest baseline.

Retrieval-augmented generation (RAG) systems depend critically on how documents are chunked and searched. Fine-grained chunks can improve retrieval precision but expand the search space, increasing latency and cost; larger chunks reduce the number of candidates but make dense similarity less reliable, as each representation mixes multiple topics and introduces more semantic noise. This trade-off becomes especially limiting in deep research tasks, where retrieval must be both fast and precise across large, heterogeneous corpora. We introduce MCompassRAG, a metadata-guided retrieval framework that uses topic-level signals as a semantic compass for selecting relevant evidence. Instead of relying only on cosine similarity between queries and noisy chunk embeddings, MCompassRAG enriches chunk representations with topic metadata defined in the same embedding space and trains a lightweight retriever through LLM-teacher distillation. At inference time, MCompassRAG performs topic-aware retrieval without additional LLM calls, improving both efficiency and evidence quality. Across six complex retrieval benchmarks, MCompassRAG improves information efficiency by 8.24% on average with over 5x lower latency than the strongest efficient RAG baselines.

Deep research (DR) systems are increasingly used for complex information-seeking tasks, but existing works mainly focus on generating reports and summaries. In contrast, many enterprise tasks instead require an agent to identify concrete workflows which is a sequence of action-steps. For example, rather than summarizing budgeting policies, an agent should be able to determine the steps needed to answer a question such as: "How do I request new headcount given a fixed budget?". Therefore, we introduce DRFLOW, a benchmark for evaluating personalized workflows predicted by agents from heterogeneous sources. Each task requires the agent to identify relevant evidence from scattered sources, then use that evidence to predict the correct action-step sequence for the user’s task. DRFLOW contains 100 tasks across five domains, with 1, 246 reference workflow steps grounded in more than 3, 900 sources. We define seven diagnostic metrics covering factual grounding, step recovery, structural ordering, condition resolution, and personalization. We further present DRFLOW-Agent (DRFA), a workflow-oriented reference agent to predict personalized workflow. We show that although DRFA improves over strong baseline agents (upto 10.02% average F1 score), there is substantial room for improvement remains across these workflow metrics, indicating that predicting complete and correct personalized workflows remains a challenging frontier for deep research.

Online web agents often augment a base actor with memory, workflow, or skill modules. These modules can improve performance, but they also consume test-time tokens, a cost rarely reported alongside the actor’s inference cost. We study online augmentation, where this overhead is paid on every task, and re-evaluate its benefits under a fixed total inference budget. We compare AWM, ASI, and ReasoningBank with a token-matched vanilla baseline that uses the same budget for additional actor steps. Across four WebArena domains and three models, Gemini 3 Flash, GPT-5.4-mini, and Qwen 3.6-27B, the vanilla baseline matches or surpasses all three augmentation methods in aggregate success rate while often using fewer total tokens. We observe a similar trend on WorkArena-L1 with Qwen 3.6-27B, indicating that the effect extends to enterprise knowledge-work tasks. Our results suggest that skills and workflow memory can be useful in specific domains, but their apparent gains often vanish against a budget-matched actor. We further show that run-to-run variance materially affects outcomes and should be reported as a core evaluation criterion for online web agents.

Existing approaches for low-resource text summarization primarily employ large language models (LLMs) like GPT-3 or GPT-4 at inference time to generate summaries directly; however, such approaches often suffer from inconsistent LLM outputs and are difficult to adapt to domain-specific data in low-resource scenarios. In this work, we propose two novel methods to effectively utilize LLMs for low-resource text summarization: 1) MixSumm, an LLM-based data augmentation regime that synthesizes high-quality documents (short and long) for few-shot text summarization, and 2) PPSL, a prompt-based pseudolabeling strategy for sample-efficient semi-supervised text summarization. Specifically, MixSumm leverages the open-source LLaMA-3-70b-Instruct model to generate new documents by mixing topical information derived from a small seed set, and PPSL leverages the LLaMA-3-70b-Instruct model to generate high-quality pseudo-labels in a semi-supervised learning setup. We evaluate our methods on the TweetSumm, WikiHow, and ArXiv/PubMed datasets and use L-Eval, a LLaMA-3-based evaluation metric, and ROUGE scores to measure the quality of generated summaries. Our experiments on extractive and abstractive summarization show that MixSumm and PPSL achieve competitive ROUGE scores as a fully supervised method with 5% of the labeled data. We release our codebase here: https://github.com/ServiceNow/text-summarization-with-llms/

Multimodal multihop question answering (MMQA) requires reasoning over images and text from multiple sources, an essential task for many real-world applications. Despite advances in visual question answering, this multihop setting remains underexplored due to a lack of quality datasets. Existing methods focus on single-hop, single-modality, or short texts, limiting real-world applications like interpreting educational documents with long, multimodal content. To fill this gap, we introduce FM2DS, the first framework for creating a high-quality dataset for MMQA. Our approach consists of a 5-stage pipeline that involves acquiring relevant multimodal documents from Wikipedia, synthetically generating high-level questions and answers, and validating them through rigorous criteria to ensure data quality. We evaluate our methodology by training models on our synthesized dataset and testing on two benchmarks: MultimodalQA and WebQA. Our results demonstrate that, with an equal sample size, models trained on our synthesized data outperform those trained on human-collected data by 1.9 in exact match (EM) score on average. Additionally, we introduce M2QA-Bench with 1k samples, the first benchmark for MMQA on long documents, generated using FM2DS and refined by human annotators.