Publications

Vision-language models (VLMs) have demonstrated impressive performance by effectively integrating visual and textual information to solve complex tasks. However, it is not clear how these models reason over the visual and textual data together, nor how the flow of information between modalities is structured. In this paper, we examine how VLMs reason by analyzing their biases when confronted with scenarios that present conflicting image and text cues, a common occurrence in real-world applications. To uncover the extent and nature of these biases, we build upon existing benchmarks to create five datasets containing mismatched image-text pairs, covering topics in mathematics, science, and visual descriptions. Our analysis shows that VLMs favor text in simpler queries but shift toward images as query complexity increases. This bias correlates with model scale, with the difference between the percentage of image- and text-preferred responses ranging from +56.8% (image favored) to -74.4% (text favored), depending on the task and model. In addition, we explore three mitigation strategies: simple prompt modifications, modifications that explicitly instruct models on how to handle conflicting information (akin to chain-of-thought prompting), and a task decomposition strategy that analyzes each modality separately before combining their results. Our findings indicate that the effectiveness of these strategies in identifying and mitigating bias varies significantly and is closely linked to the model’s overall performance on the task and the specific modality in question.

Retrieval-augmented generation (RAG) and long-context language models (LCLMs) both address context limitations of LLMs in open-domain question answering (QA). However, optimal external context to retrieve remains an open problem: fixing the retrieval size risks either wasting tokens or omitting key evidence. Existing adaptive methods like Self-RAG and Self-Route rely on iterative LLM prompting and perform well on factoid QA, but struggle with aggregation QA, where the optimal context size is both unknown and variable. We present Adaptive-k retrieval, a simple and effective single-pass method that adaptively selects the number of passages based on the distribution of the similarity scores between the query and the candidate passages. It does not require model fine-tuning, extra LLM inferences or changes to existing retriever-reader pipelines. On both factoid and aggregation QA benchmarks, Adaptive-k matches or outperforms fixed-k baselines while using up to 10x fewer tokens than full-context input, yet still retrieves 70% of relevant passages. It improves accuracy across five LCLMs and two embedding models, highlighting that dynamically adjusting context size leads to more efficient and accurate QA.

Recent advances in large language models (LLMs) have enabled impressive performance in various tasks. However, standard prompting often struggles to produce structurally valid and accurate outputs, especially in dependency parsing. We propose a novel step-by-step instruction strategy, where universal part-of-speech tagging precedes the prediction of syntactic heads and dependency labels, and a simplified CoNLL-U like output format, our method achieves state-of-the-art accuracy on Universal Dependencies datasets across 17 languages without hallucination or contamination. We further show that multilingual fine-tuning simultaneously improves cross-language generalization performance. Our results highlight the effectiveness of explicit reasoning steps in LLM-based parsing and offer a scalable, format-consistent alternative to bracket-based approaches.

NL2SQL approaches have greatly benefited from the impressive capabilities of large language models (LLMs). In particular, bootstrapping an NL2SQL system for a specific domain can be as simple as instructing an LLM with sufficient contextual information, such as schema details and translation demonstrations. However, building an accurate system still requires the rigorous task of selecting the right context for each query-including identifying relevant schema elements, cell values, and suitable exemplars that help the LLM understand domain-specific nuances. Retrieval-based methods have become the go-to approach for identifying such context. While effective, these methods introduce additional inference-time costs due to the retrieval process. In this paper, we argue that production scenarios demand high-precision, high-performance NL2SQL systems, rather than simply high-quality SQL generation, which is the focus of most current NL2SQL approaches. In such scenarios, the careful selection of a static set of exemplars-capturing the intricacies of the query log, target database, SQL constructs, and execution latencies-plays a more crucial role than exemplar selection based solely on similarity. The key challenge, however, lies in identifying a representative set of exemplars for a given production setting. To this end, we propose a prompt optimization framework that not only addresses the high-precision requirement but also optimizes the performance of the generated SQL through multi-objective optimization. Preliminary empirical analysis demonstrates the effectiveness of the proposed framework.

近年，企業の製品やサービスに対する問い合わせ件数は増加の一途をたどっており，人手による個別対応では非効率性が生じるだけでなく，迅速な回答が得られないことによる顧客満足度の低下が課題となっている．これに対処する手段として，チャットボットを活用した自動化が注目されているが，その精度や有効性を向上させるためには，ユーザーが抱える問題の背景や意図を正確に把握する必要がある．本研究では，ユーザー問い合わせと関連情報を体系的に構造化するナレッジグラフの生成に焦点を当てる．具体的には，多様な問い合わせに含まれる曖昧表現や不完全な情報を扱うために，FAQ をはじめとしたヘルプページから，意図推定や回答文生成に資するナレッジグラフの構築を行う．本手法で構築されたナレッジグラフを用いることで，将来的にチャットボット等の自動対応システムを用いた，大量問い合わせへの効率的な対応を可能になると考えられる．

近年，構文解析分野でも大規模言語モデル (LLM) の応用が進んでいる．本論文では，事後学習により対話応答性能が強化された LLM に対して，入力文とその依存構造解析結果の関係を指示-応答のプロンプト形式で表現した教師データで Fine-tuning を行い，指示プロンプトにより依存構造解析タスクを実行する対話型依存構造解析を提案する．提案手法は，先行研究で用いられてきた括弧の対応関係による構文木表現は用いず，単語テーブル上で依存先の単語インデックスを参照する形で依存構造を表現することで，可読性を向上すると同時に，交差を含めた依存構造を表現可能にしている．提案手法の解析精度をUniversal Dependencies の英語と日本語のデータセットで評価した結果，英語においては提案手法が従来のパイプライン型の解析手法の精度を上回り，日本語においては同程度の精度となることを確認したが，ベースモデルの学習時に評価用データセットが混入している可能性を考慮する必要がある．同時に，提案手法の出力は非巡回・シングルルートなど一般的な構文木の制約を満たさない場合がごく稀にあること，ハルシネーションにより系列が部分的に変化し入力と出力と整合しない場合があることを確認した．提案手法は一般的な LLM ホスティングサービス上で容易に実現できるため，従来のパイプライン型の解析処理に比べて開発負荷を大きく減らすことが可能であり，今後の利用の広まりが期待される．

Large Language Models (LLMs) have shown impressive capability in language generation and understanding, but their tendency to hallucinate and produce factually incorrect information remains a key limitation. To verify LLM-generated contents and claims from other sources, traditional verification approaches often rely on holistic models that assign a single factuality label to complex claims, potentially obscuring nuanced errors. In this paper, we advocate for a shift toward fine-grained verification, where complex claims are broken down into smaller sub-claims for individual verification, allowing for more precise identification of inaccuracies, improved transparency, and reduced ambiguity in evidence retrieval. However, generating sub-claims poses challenges, such as maintaining context and ensuring semantic equivalence with respect to the original claim. We introduce FactLens, a benchmark for evaluating fine-grained fact verification, with metrics and automated evaluators of sub-claim quality. The benchmark data is manually curated to ensure high-quality ground truth. Our results show alignment between automated FactLens evaluators and human judgments, and we discuss the impact of sub-claim characteristics on the overall verification performance.

Retrieval from graph data is crucial for augmenting large language models (LLM) with both open-domain knowledge and private enterprise data, and it is also a key component in the recent GraphRAG system (edge et al., 2024). Despite decades of research on knowledge graphs and knowledge base question answering, leading LLM frameworks (e.g. Langchain and LlamaIndex) have only minimal support for retrieval from modern encyclopedic knowledge graphs like Wikidata. In this paper, we analyze the root cause and suggest that modern RDF knowledge graphs (e.g. Wikidata, Freebase) are less efficient for LLMs due to overly large schemas that far exceed the typical LLM context window, use of resource identifiers, overlapping relation types and lack of normalization. As a solution, we propose property graph views on top of the underlying RDF graph that can be efficiently queried by LLMs using Cypher. We instantiated this idea on Wikidata and introduced CypherBench, the first benchmark with 11 large-scale, multi-domain property graphs with 7.8 million entities and over 10,000 questions. To achieve this, we tackled several key challenges, including developing an RDF-to-property graph conversion engine, creating a systematic pipeline for text-to-Cypher task generation, and designing new evaluation metrics.

Large Language Models (LLMs) offer the potential to automate hiring by matching job descriptions with candidate resumes, streamlining recruitment processes, and reducing operational costs. However, biases inherent in these models may lead to unfair hiring practices, reinforcing societal prejudices and undermining workplace diversity. This study examines the performance and fairness of LLMs in job-resume matching tasks within the English language and U.S. context. It evaluates how factors such as gender, race, and educational background influence model decisions, providing critical insights into the fairness and reliability of LLMs in HR applications. Our findings indicate that while recent models have reduced biases related to explicit attributes like gender and race, implicit biases concerning educational background remain significant. These results highlight the need for ongoing evaluation and the development of advanced bias mitigation strategies to ensure equitable hiring practices when using LLMs in industry settings.

Utilizing large language models (LLMs) to rank a set of items has become a common approach in recommendation and retrieval systems. Typically, these systems focus on ordering a substantial number of documents in a monotonic order based on a given query. However, real-world scenarios often present a different challenge: ranking a comparatively smaller set of items, but according to a variety of diverse and occasionally conflicting conditions. In this paper, we define and explore the task of multi-conditional ranking by introducing MCRank, a benchmark tailored for assessing multi-conditional ranking across various item types and conditions. Our analysis of LLMs using MCRank indicates a significant decrease in performance as the number and complexity of items and conditions grow. To overcome this limitation, we propose a novel decomposed reasoning method, consisting of EXtracting and Sorting the conditions, and then Iteratively Ranking the items (EXSIR). Our extensive experiments show that this decomposed reasoning method enhances LLMs’ performance significantly, achieving up to a 14.4% improvement over existing LLMs. We also provide a detailed analysis of LLMs performance across various condition categories, and examine the effectiveness of decomposition step. Furthermore, we compare our method with existing approaches such as Chain-of-Thought and existing ranking models, demonstrating the superiority of our approach and complexity of MCR task. We released our dataset and code.