Reproducibility must validate architectural robustness, not just numerical accuracy. We evaluate ColBERT-v2 and ConstBERT across five dimensions, finding that while ConstBERT reproduces within 0.05% MRR@10 on MS-MARCO, both models show a drop of 86–97% on long, narrative queries (TREC ToT 2025). Ablations prove this failure is architectural: performance plateaus at 20 words because the MaxSim operator's uniform token weighting cannot distinguish signal from filler noise. Furthermore, undocumented backend parameters create an 8-point gap due to ConstBERT's sparse centroid coverage, and fine-tuning with 3× more data actually degrades performance by up to 29%. We conclude that architectural constraints in multi-vector retrieval cannot be overcome by adaptation alone. Code: https://github.com/utshabkg/multi-vector-reproducibility.

Large Language Models (LLMs) have transformed information access by enabling human-like text understanding and generation. This workshop explores the next step for conversational AI: building proactive information-seeking assistants that go beyond reactive question answering. We aim to investigate how LLMs can anticipate user needs, model complex context, support mixed-initiative interactions, integrate retrieval and external tools, personalize responses, adapt through feedback, and ensure fairness, transparency, and cognitive grounding. Bringing together experts from NLP, IR, HCI, and cognitive science, the workshop will serve as a timely forum for advancing intelligent, proactive dialogue systems. It will also foster interdisciplinary collaboration.

Neural IR has advanced through two distinct paths: entity-oriented approaches leveraging knowledge graphs and multi-vector models capturing fine-grained semantics. We introduce QDER, a neural re-ranking model that unifies these approaches by integrating knowledge graph semantics into a multi-vector model.

Conversational information seeking has evolved rapidly in the last few years with the development of Large Language Models (LLMs), providing the basis for interpreting and responding in a naturalistic manner to user requests. The extended TREC Interactive Knowledge Assistance Track (iKAT) collection aims to enable researchers to test and evaluate their Conversational Search Agent (CSA). The collection contains a set of 36 personalized dialogues over 20 different topics each coupled with a Personal Text Knowledge Base (PTKB) that defines the bespoke user personas. A total of 344 turns with approximately 26,000 passages are provided as assessments on relevance, as well as additional assessments on generated responses over four key dimensions: relevance, completeness, groundedness, and naturalness. The collection challenges CSAs to efficiently navigate diverse personal contexts, elicit pertinent persona information, and employ context for relevant conversations. The integration of a PTKB and the emphasis on decisional search tasks contribute to the uniqueness of this test collection, making it an essential benchmark for advancing research in conversational and interactive knowledge assistants.

Contains fulltext : 308793.pdf (Publisher’s version ) (Open Access)

This tutorial will provide an overview of recent advances on neuro-symbolic approaches for information retrieval. A decade ago, knowledge graphs and semantic annotations technology led to active research on how to best leverage symbolic knowledge. At the same time, neural methods have demonstrated to be versatile and highly effective.

Current query expansion models use pseudo-relevance feedback to improve first-pass retrieval effectiveness; however, this fails when the initial results are not relevant. Instead of building a language model from retrieved results, we propose Generative Relevance Feedback (GRF) that builds probabilistic feedback models from long-form text generated from Large Language Models. We study the effective methods for generating text by varying the zero-shot generation subtasks: queries, entities, facts, news articles, documents, and essays. We evaluate GRF on document retrieval benchmarks covering a diverse set of queries and document collections, and the results show that GRF methods significantly outperform previous PRF methods. Specifically, we improve MAP between 5-19% and NDCG@10 17-24% compared to RM3 expansion, and achieve state-of-the-art recall across all datasets.

Entity linking can disambiguate mentions of an entity in text. However, there are many different aspects of an entity that could be discussed but are not differentiable by entity links, for example, the entity "oyster'' in the context of "food'' or "ecosystems''. Entity aspect linking provides such fine-grained explicit semantics for entity links by identifying the most relevant aspect of an entity in the given context. We propose a novel entity aspect linking approach that outperforms several neural and non-neural baselines on a large-scale entity aspect linking test collection. Our approach uses a supervised neural entity ranking system to predict relevant entities for the context. These entities are then used to guide the system to the correct aspect.

Entity-oriented search systems often learn vector representations of entities via the introductory paragraph from the Wikipedia page of the entity. As such representations are the same for every query, our hypothesis is that the representations are not ideal for IR tasks. In this work, we present BERT Entity Representations (BERT-ER) which are query-specific vector representations of entities obtained from text that describes how an entity is relevant for a query. Using BERT-ER in a downstream entity ranking system, we achieve a performance improvement of 13-42% (Mean Average Precision) over a system that uses the BERT embedding of the introductory paragraph from Wikipedia on two large-scale test collections. Our approach also outperforms entity ranking systems using entity embeddings from Wikipedia2Vec, ERNIE, and E-BERT. We show that our entity ranking system using BERT-ER can increase precision at the top of the ranking by promoting relevant entities to the top. With this work, we release our BERT models and query-specific entity embeddings fine-tuned for the entity ranking task.

We provide a resource for automatically harvesting relevance benchmarks from Wikipedia -- which we refer to as "Wikimarks" to differentiate them from manually created benchmarks. Unlike simulated benchmarks, they are based on manual annotations of Wikipedia authors. Studies on the TREC Complex Answer Retrieval track demonstrated that leaderboards under Wikimarks and manually annotated benchmarks are very similar. Because of their availability, Wikimarks can fill an important need for Information Retrieval research.

Using entity aspect links, we improve upon the current state-of-the-art in entity retrieval. Entity retrieval is the task of retrieving relevant entities for search queries, such as "Antibiotic Use In Livestock". Entity aspect linking is a new technique to refine the semantic information of entity links. For example, while passages relevant to the query above may mention the entity "USA", there are many aspects of the USA of which only few, such as "USA/Agriculture", are relevant for this query. By using entity aspect links that indicate which aspect of an entity is being referred to in the context of the query, we obtain more specific relevance indicators for entities. We show that our approach improves upon all baseline methods, including the current state-of-the-art using a standard entity retrieval test collection. With this work, we release a large collection of entity-aspect-links for a large TREC corpus.