--- title: "Rui Wang — papers, page 2 — Elamur Research" description: "Rui Wang, Huaqiao University: 33 papers on Elamur Research (2007–2026), h-index 37 — affiliations, topics, co-authors and who cites them…" canonical: "https://research.elamur.ai/free/en/authors/rui-wang/page/2/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/rui-wang/page/2/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / [Rui Wang](https://md.research.elamur.ai/free/en/authors/rui-wang/) / page 2 # Rui Wang · page 2 33 papers in our library · 2007–2026 · page 2 of 2 [The profile, affiliations and co-authors →](https://md.research.elamur.ai/free/en/authors/rui-wang/) ## Papers The most cited first (OpenAlex and our library). [← prev](https://md.research.elamur.ai/free/en/authors/rui-wang/) [1](https://md.research.elamur.ai/free/en/authors/rui-wang/) **2** 26. **[Robust Tool Use via Fission-GRPO: Learning to Recover from Execution Errors](https://md.research.elamur.ai/free/en/papers/zhang-2026-robust-tool-use-fission-grpo/)** *(2026 · 1 citation)* Small tool-calling LLMs handle multi-turn tasks poorly once a tool call fails. Instead of reading the error message and adjusting, they often repeat the same invalid call until the dialogue breaks down. On BFCL v4 Multi-Turn, Qwen3-8B recovers in only about 20% of such cases, while Claude Sonnet 4 recovers in over… 27. **[WebAggregator: Enhancing Compositional Reasoning Capabilities of Deep Research Agent Foundation Models](https://md.research.elamur.ai/free/en/papers/wang-2026-webaggregator-enhancing-compositional-reasoning-capabilities/)** *(2026 · 1 citation)* Deep Research agents need more than retrieval. They must combine scattered, heterogeneous evidence through multi-step reasoning. Existing training and evaluation resources mostly reward locating a single entity, so agents learn to call tools often instead of reasoning carefully. 28. **[BabelDOC: Better Layout-Preserving PDF Translation via Intermediate Representation](https://md.research.elamur.ai/free/en/papers/qi-2026-babeldoc-better-layout-preserving-pdf/)** *(2026)* PDFs are built for faithful display rather than editing, which makes translating them difficult. Text-focused MT and CAT tools lose structural information. Parsers such as Doc2X and MinerU extract content in one direction only and cannot rebuild the original page. BabelDOC tackles this with a bidirectional… 29. **[CachePrune: Teaching LLMs What Not to Follow via KV-Cache Editing](https://md.research.elamur.ai/free/en/papers/wang-2026-cacheprune-teaching-llms-what-not/)** *(2026)* The paper addresses indirect prompt injection, where an LLM obeys commands hidden inside supplied context (e.g., a retrieved passage) instead of answering the user's actual request. 30. **[Mitigating Context Interference for Reliable and Efficient Search Agents](https://md.research.elamur.ai/free/en/papers/xue-2026-mitigating-context-interference-reliable-efficient/)** *(2026)* Multi-turn LLM search agents repeatedly query a retriever and add the returned documents to their working context. 31. **[RepoDistill: Distilling Repository Knowledge through Compression-Aware Budget Allocation and Policy Optimization](https://md.research.elamur.ai/free/en/papers/yin-2026-repodistill-distilling-repository-knowledge-through/)** *(2026)* Repository-level coding tasks require LLMs to work with long, noisy contexts that span many files. Retrieving more code often buries the relevant parts and lowers accuracy. In the authors' pilot study with function-chunking RAG, scores rose until about 30k tokens of context and then fell sharply at 100k–200k tokens… 32. **[Self-Sum: Teaching an Agent to Decide Itself When and What to Summarize](https://md.research.elamur.ai/free/en/papers/wang-2026-self-sum-teaching-agent-decide/)** *(2026)* Agents that tackle long tasks pile up reasoning steps, actions and observations. Their context grows noisy and expensive to process. Common fixes compress history on a fixed schedule or hand the job to an external model. These approaches ignore where the agent actually is in the task and can discard crucial details.… 33. **[TingIS: Real-time Risk Event Discovery from Noisy Customer Incidents at Enterprise Scale](https://md.research.elamur.ai/free/en/papers/wang-2026-tingis-real-time-risk-event/)** *(2026)* Large cloud-native and fintech services sometimes fail in ways that internal monitoring does not catch. Complaints from users can expose these failures early, but the complaints are colloquial, noisy, arrive in heavy streams and span many business lines. TingIS is an end-to-end system from Ant Group that turns this… [← prev](https://md.research.elamur.ai/free/en/authors/rui-wang/) [1](https://md.research.elamur.ai/free/en/authors/rui-wang/) **2**