Conversational search agents, especially with the advent of large language models, have developed into useful tools to satisfy complex information needs of their users. Former research has shown that personalization (i.e., adaptation of agent responses to the preferences and traits of the user) can increase the relevance and perceived answer quality of these systems even further. However, developing accurate personalization methods typically requires rich datasets, both in terms of user profiles and complex conversations, for which only a few resources are publicly available. Over the past three years, the goal of the TREC Interactive Knowledge Assistance Track (iKAT) has been to bridge this gap. In organizing this shared task, we have developed a collection of complex information needs and associated conversations, made to challenge today's conversational agents and thus highlight aspects in need of further research. In this paper, we present the resources made for iKAT 2025, focusing on multi-session conversations (i.e., multiple dialogues per user), dynamically evolving user models, mixed-initiative dialogues, and large-scale human and automatic assessments. In addition to manually designed user profiles and conversations, the test collection for 2025 also contains dialogues between participating systems and our user simulators. All the resources are publicly available in our repository, including the system evaluation both as a result of the offline (i.e., test collection-based) and interactive tasks (i.e., user simulation-based), as well as their source code and model weights, to foster future research in this direction.

Generative retrieval (GR) ranks documents by autoregressively generating document identifiers. Because many GR methods rely on trie-constrained beam search, they are vulnerable to early pruning of relevant prefixes under finite-beam decoding. Planning Ahead in Generative Retrieval (PAG) mitigates this failure mode by using simultaneous decoding to compute a document-level look-ahead prior that guides subsequent sequential decoding. We reproduce PAG at inference time and stress-test its decoding behavior. Using the authors' released checkpoint and identifier/trie artifacts under the reported decoding setup, we reproduce the main effectiveness results on MS~MARCO Dev and TREC-DL 2019/2020, and corroborate the reported beam-size--latency trade-off in our hardware setting. Beyond reproduction, we introduce plan drift diagnostics that quantify how intent-preserving query variations, including misspellings, reordering, synonym substitutions, paraphrases, and naturality shifts, alter the planner's top-n candidate set and highest-weight planner tokens, and how these changes affect guided decoding. We find that PAG's planning signal is brittle under lexical surface-form variation: intent-preserving typos can trigger plan collapse, where the planned candidate pool shifts enough that the look-ahead bonus provides little useful guidance, effectively reverting decoding toward weaker unguided search. We further evaluate fixed-index cross-lingual robustness using non-English mMARCO queries against an English index, and assess query-side mitigation strategies that require no re-indexing; query translation provides the strongest recovery in our setting. Overall, our results confirm PAG's reported effectiveness and the benefit of planning-guided decoding under the released inference setup, while showing that these gains depend on the stability of the planning signal under realistic query variation and query--document mismatch. Code available at https://github.com/kidist-amde/lost-in-decoding.

This paper concerns corpus poisoning attacks in dense information retrieval, where an adversary attempts to compromise the ranking performance of a search algorithm by injecting a small number of maliciously generated documents into the corpus.Our work addresses two limitations in the current literature.First, attacks that perform adversarial gradient-based word substitution search do so in the discrete lexical space, while retrieval itself happens in the continuous embedding space.We thus propose an optimization method that operates in the embedding space directly.Specifically, we train a perturbation model with the objective of maintaining the geometric distance between the original and adversarial document embeddings, while also maximizing the token-level dissimilarity between the original and adversarial documents.Second, it is common for related work to have a strong assumption that the adversary has prior knowledge about the queries.In this paper, we focus on a more challenging variant of the problem where the adversary assumes no prior knowledge about the query distribution (hence, unsupervised).Our core contribution is an adversarial corpus attack that is fast and effective.We present comprehensive experimental results on both in-and out-of-domain datasets, focusing on two related tasks: a top-1 attack and a corpus poisoning attack.We consider attacks under both a white-box and a black-box setting.Notably, our method can generate successful adversarial examples in under two minutes per target document; four times faster compared to the fastest gradientbased word substitution methods in the literature with the same hardware.Furthermore, our adversarial generation method generates text that is more likely to occur under the distribution of natural text (low perplexity), and is therefore more difficult to detect.

The rise of personalized conversational search systems has been driven by advancements in Large Language Models (LLMs), enabling these systems to retrieve and generate answers for complex information needs. However, the automatic evaluation of responses generated by Retrieval Augmented Generation (RAG) systems remains an understudied challenge. In this paper, we introduce a new resource for assessing the retrieval effectiveness and relevance of responses generated by RAG systems, using a nugget-based evaluation framework. Built upon the foundation of TREC iKAT 2023, our dataset extends to the TREC iKAT 2024 collection, which includes 17 conversations and 20,575 relevance passage assessments, together with 2,279 extracted gold nuggets and 62 manually written gold answers from NIST assessors. While maintaining the core structure of its predecessor, this new collection enables a deeper exploration of generation tasks in conversational settings. Key improvements in iKAT 2024 include: (1) ''gold nuggets'' - concise, essential pieces of information extracted from relevant passages of the collection - which serve as a foundation for automatic response evaluation; (2) manually written answers to provide a gold standard for response evaluation; (3) expanded user personas, providing richer contextual grounding; and (4) a transition from Personal Text Knowledge Base (PTKB) ranking to PTKB classification and selection. Built on this resource, we provide a framework for long-form answer generation evaluation, involving nugget extraction and nugget matching, linked to retrieval. This establishes a solid resource for advancing research in personalized conversational search and long-form answer generation. Our resources are publicly available at https://github.com/irlabamsterdam/CONE-RAG.

Negation is a fundamental aspect of human communication, yet it remains a challenge for Language Models (LMs) in Information Retrieval (IR). Despite the heavy reliance of modern neural IR systems on LMs, little attention has been given to their handling of negation. In this study, we reproduce and extend the findings of NevIR, a benchmark study that revealed most IR models perform at or below the level of random ranking when dealing with negation. We replicate NevIR's original experiments and evaluate newly developed state-of-the-art IR models. Our findings show that a recently emerging category-listwise Large Language Model (LLM) re-rankers-outperforms other models but still underperforms human performance. Additionally, we leverage ExcluIR, a benchmark dataset designed for exclusionary queries with extensive negation, to assess the generalisability of negation understanding. Our findings suggest that fine-tuning on one dataset does not reliably improve performance on the other, indicating notable differences in their data distributions. Furthermore, we observe that only cross-encoders and listwise LLM re-rankers achieve reasonable performance across both negation tasks.

Conversational Search (CS) involves retrieving relevant documents from a corpus while considering the conversational context, integrating retrieval with context modeling. Recent advancements in Large Language Models (LLMs) have significantly enhanced CS by enabling query rewriting based on conversational context. However, employing LLMs during inference poses efficiency challenges. Existing solutions mitigate this issue by distilling embeddings derived from human-rewritten queries, focusing primarily on learning the context modeling task. These methods, however, often separate the contrastive retrieval task from the distillation process, treating it as an independent loss term. To overcome these limitations, we introduce DiSCo (Distillation of Sparse Conversational retrieval), a novel approach that unifies retrieval and context modeling through a relaxed distillation objective. Instead of relying exclusively on representation learning, our method distills similarity scores between conversations and documents, providing more freedom in the representation space and better leveraging the contrastive nature of document relevance. Extensive experiments on Learned Sparse Retrieval (LSR) across five CS datasets demonstrate that DiSCo achieves substantial improvements in both in-domain and out-of-domain retrieval tasks, achieving up to a six-point gain in recall for out-of-domain datasets over state-of-the-art methods. Additionally, DiSCo employs a multi-teacher distillation strategy, using multiple LLMs as teachers, further enhancing performance and surpassing the individual teachers in in-domain settings. Furthermore, analysis of model sparsity reveals that DiSCo allows for more effective control over the sparsity of the trained models.

Middle training methods aim to bridge the gap between the Masked Language Model (MLM) pre-training and the final finetuning for retrieval. Recent models such as CoCondenser, RetroMAE, and LexMAE argue that the MLM task is not sufficient enough to pre-train a transformer network for retrieval and hence propose various tasks to do so. Intrigued by those novel methods, we noticed that all these models used different finetuning protocols, making it hard to assess the benefits of middle training. We propose in this paper a benchmark of CoCondenser, RetroMAE, and LexMAE, under the same finetuning conditions. We compare both dense and sparse approaches under various finetuning protocols and middle training on different collections (MS MARCO, Wikipedia). We use additional middle training baselines, such as a standard MLM finetuning on the retrieval collection, optionally augmented by a CLS predicting the passage term frequency. For the sparse approach, our study reveals that there is almost no statistical difference between those methods: the more effective the finetuning procedure is, the less difference there is between those models. For the dense approach, RetroMAE using MS MARCO as middle-training collection shows excellent results in almost all the settings. Finally, we show that middle training on the retrieval collection, thus adapting the language model to it, is a critical factor. Overall, a better experimental setup should be adopted to evaluate middle training methods.

Sparse neural retrievers, such as DeepImpact, uniCOIL and SPLADE, have been introduced recently as an efficient and effective way to perform retrieval with inverted indexes. They aim to learn term importance and, in some cases, document expansions, to provide a more effective document ranking compared to traditional bag-of-words retrieval models such as BM25. However, these sparse neural retrievers have been shown to increase the computational costs and latency of query processing compared to their classical counterparts. To mitigate this, we apply a well-known family of techniques for boosting the efficiency of query processing over inverted indexes: static pruning. We experiment with three static pruning strategies, namely document-centric, term-centric and agnostic pruning, and we assess, over diverse datasets, that these techniques still work with sparse neural retrievers. In particular, static pruning achieves 2x speedup with negligible effectiveness loss (≤ 2% drop) and, depending on the use case, even 4x speedup with minimal impact on the effectiveness (≤ 8% drop). Moreover, we show that neural rerankers are robust to candidates from statically pruned indexes.