Sep 2026· Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing· Vol abs/2609.21247· 0 citations· 30 references
Computer Science
TL;DR
HMS is introduced, a scalable framework that induces coarse- and fine-grained reasoning structures without predefined taxonomies and reveals a shared organization--understanding/planning, translating/drafting, and refining/verifying--alongside domain-specific variation.
Abstract
Large Reasoning Models increasingly use intermediate traces for machine translation, but it remains unclear when such reasoning helps or hurts. We analyze reasoning traces across models, languages, domains, and datasets, focusing on reasoning language, length, and structure. We find that the best reasoning language is model-specific, reasoning length has a non-monotonic relationship with quality, and traces exhibit recurring functional patterns. To uncover these patterns, we introduce Hierarchical Meta-Summarization (HMS), a scalable framework that induces coarse- and fine-grained reasoning structures without predefined taxonomies. HMS reveals a shared organization--understanding/planning, translating/drafting, and refining/verifying--alongside domain-specific variation. Our results suggest that MT reasoning should be controlled in a model-aware, length-aware, and pattern-aware manner rather than uniformly encouraged.
A novel fragment-based reasoning framework is introduced in which the model first extracts parallel source-target fragments from retrieved similar exemplars, and uses these fragments as intermediate reasoning traces to produce the final translation.
Maxime Bouthors, J. Crego, François Yvon· 0 citations
Large Language Models (LLMs) achieve strong performance on many tasks, yet their reasoning can remain brittle and their decision process is often insufficiently transparent. Knowledge graphs (KGs) provide explicit entities and relations that can ground reasoning and improve traceability. However, most KG-augmented appr...
Jia-Jun Duan, Qiu-Ling Chen, Ze-Xu Wang et al.· 2026 6th International Confe...· 0 citations
This work builds a fully monolingual German RAG question-answering testbed over the fictional world of the tabletop role-playing game The Dark Eye, a domain that is richly documented in German but too niche for the model to answer from memory, so that it has to rely on retrieval.
Oliver Hauck, Mario Sanz-Guerrero, Katharina von der Wense· The Proceedings of the Works...· 0 citations
GRS-QA is introduced, a reasoning graph-structured question answering dataset that augments multi-hop QA instances from exisiting datasets, with directed reasoning graphs representing intermediate inference steps, and highlights the importance of unified structural annotations for understanding and improving multi-hop...
Khin S. Yone, Devasha Trivedi, Anish Pahilajani et al.· International Conference on...· 0 citations
Skeleton-based reasoning prompting is a promising training-free approach for structuring LLM reasoning, but prior work largely assumes an English-centric setting. We propose the Language-Aware Skeleton Exploration Framework (LASEF) to study skeleton-language choice in multilingual mathematical reasoning. Across math be...
HyeonSeok Lim, Seung-Woo Song, Inho Won et al.· Conference on Empirical Meth...· 0 citations
Logical reasoning with large language models (LLMs) is a critical capability, as it reflects a system's ability to correctly deduce hypotheses from a given context using faithful deductive processes. However, LLM reasoning has often been shown to be sensitive to small surface-level variations in problem formulation, ra...
Ramya Keerthy Thatikonda, W. Buntine, Ehsan Shareghi· 0 citations
Related blog posts
MIT News · Artificial Intelligence· news.mit.eduNov 19, 2026
An MIT and Dana-Farber Institute-led team used large language models to examine nearly 4,000 studies and hundreds of code repositories, revealing gaps in the information needed to scrutinize and reuse published research.
By studying the computational foundations of language learning, Associate Professor Jacob Andreas builds intelligent systems that can communicate effectively with humans.