Akari Asai

Sheng-Chieh Lin, Akari Asai, Minghan Li et al. · meta-ai, uw

Various techniques have been developed in recent years to improve dense retrieval (DR), such as unsupervised contrastive learning and pseudo-query generation. Existing DRs, however, often suffer from effectiveness tradeoffs between supervised and zero-shot retrieval, which some argue was due to the limited model capacity. We contradict this hypothesis and show that a generalizable DR can be trained to achieve high accuracy in both supervised and zero-shot retrieval without increasing model size. In particular, we systematically examine the contrastive learning of DRs, under the framework of Data Augmentation (DA). Our study shows that common DA practices such as query augmentation with generative models and pseudo-relevance label creation using a cross-encoder, are often inefficient and sub-optimal. We hence propose a new DA approach with diverse queries and sources of supervision to progressively train a generalizable DR. As a result, DRAGON, our dense retriever trained with diverse augmentation, is the first BERT-base-sized DR to achieve state-of-the-art effectiveness in both supervised and zero-shot evaluations and even competes with models using more complex late interaction (ColBERTv2 and SPLADE++).

Akari Asai, Zeqiu Wu, Yizhong Wang et al. · uw

Despite their remarkable capabilities, large language models (LLMs) often produce responses containing factual inaccuracies due to their sole reliance on the parametric knowledge they encapsulate. Retrieval-Augmented Generation (RAG), an ad hoc approach that augments LMs with retrieval of relevant knowledge, decreases such issues. However, indiscriminately retrieving and incorporating a fixed number of retrieved passages, regardless of whether retrieval is necessary, or passages are relevant, diminishes LM versatility or can lead to unhelpful response generation. We introduce a new framework called Self-Reflective Retrieval-Augmented Generation (Self-RAG) that enhances an LM's quality and factuality through retrieval and self-reflection. Our framework trains a single arbitrary LM that adaptively retrieves passages on-demand, and generates and reflects on retrieved passages and its own generations using special tokens, called reflection tokens. Generating reflection tokens makes the LM controllable during the inference phase, enabling it to tailor its behavior to diverse task requirements. Experiments show that Self-RAG (7B and 13B parameters) significantly outperforms state-of-the-art LLMs and retrieval-augmented models on a diverse set of tasks. Specifically, Self-RAG outperforms ChatGPT and retrieval-augmented Llama2-chat on Open-domain QA, reasoning and fact verification tasks, and it shows significant gains in improving factuality and citation accuracy for long-form generations relative to these models.

Alex Mallen, Akari Asai, Victor Zhong et al. · uw

Despite their impressive performance on diverse tasks, large language models (LMs) still struggle with tasks requiring rich world knowledge, implying the limitations of relying solely on their parameters to encode a wealth of world knowledge. This paper aims to understand LMs' strengths and limitations in memorizing factual knowledge, by conducting large-scale knowledge probing experiments of 10 models and 4 augmentation methods on PopQA, our new open-domain QA dataset with 14k questions. We find that LMs struggle with less popular factual knowledge, and that scaling fails to appreciably improve memorization of factual knowledge in the long tail. We then show that retrieval-augmented LMs largely outperform orders of magnitude larger LMs, while unassisted LMs remain competitive in questions about high-popularity entities. Based on those findings, we devise a simple, yet effective, method for powerful and efficient retrieval-augmented LMs, which retrieves non-parametric memories only when necessary. Experimental results show that this significantly improves models' performance while reducing the inference costs.

Akari Asai, Timo Schick, Patrick Lewis et al. · meta-ai, uw

We study the problem of retrieval with instructions, where users of a retrieval system explicitly describe their intent along with their queries. We aim to develop a general-purpose task-aware retrieval system using multi-task instruction tuning, which can follow human-written instructions to find the best documents for a given query. We introduce the first large-scale collection of approximately 40 retrieval datasets with instructions, BERRI, and present TART, a multi-task retrieval system trained on BERRI with instructions. TART shows strong capabilities to adapt to a new retrieval task via instructions and advances the state of the art on two zero-shot retrieval benchmarks, BEIR and LOTTE, outperforming models up to three times larger. We further introduce a new evaluation setup, X^2-Retrieval to better reflect real-world scenarios, where diverse domains and tasks are pooled and a system needs to find documents aligning users' intents. In this setup, TART significantly outperforms competitive baselines, further demonstrating the effectiveness of guiding retrieval with instructions.

Jungo Kasai, Keisuke Sakaguchi, Yoichi Takahashi et al. · allen-ai, uw

We introduce REALTIME QA, a dynamic question answering (QA) platform that announces questions and evaluates systems on a regular basis (weekly in this version). REALTIME QA inquires about the current world, and QA systems need to answer questions about novel events or information. It therefore challenges static, conventional assumptions in open-domain QA datasets and pursues instantaneous applications. We build strong baseline models upon large pretrained language models, including GPT-3 and T5. Our benchmark is an ongoing effort, and this paper presents real-time evaluation results over the past year. Our experimental results show that GPT-3 can often properly update its generation results, based on newly-retrieved documents, highlighting the importance of up-to-date information retrieval. Nonetheless, we find that GPT-3 tends to return outdated answers when retrieved documents do not provide sufficient information to find an answer. This suggests an important avenue for future research: can an open-domain QA system identify such unanswerable cases and communicate with the user or even the retrieval module to modify the retrieval results? We hope that REALTIME QA will spur progress in instantaneous applications of question answering and beyond.

Akari Asai, Mohammadreza Salehi, Matthew E. Peters et al. · uw

This work introduces a new multi-task, parameter-efficient language model (LM) tuning method that learns to transfer knowledge across different tasks via a mixture of soft prompts-small prefix embedding vectors pre-trained for different tasks. Our method, called ATTEMPT (ATTEntional Mixtures of Prompt Tuning), obtains source prompts as encodings of large-scale source tasks into a small number of parameters and trains an attention module to interpolate the source prompts and a newly initialized target prompt for every instance in the target task. During training, only the target task prompt and the attention weights, which are shared between tasks in multi-task training, are updated, while the original LM and source prompts are intact. ATTEMPT is highly parameter-efficient (e.g., updates 2,300 times fewer parameters than full fine-tuning) while achieving high task performance using knowledge from high-resource tasks. Moreover, it is modular using pre-trained soft prompts, and can flexibly add or remove source prompts for effective knowledge transfer. Our experimental results across 21 diverse NLP datasets show that ATTEMPT significantly outperforms prompt tuning and outperforms or matches fully fine-tuned or other parameter-efficient tuning approaches that use over ten times more parameters. Finally, ATTEMPT outperforms previous work in few-shot learning settings.

Rulin Shao, Jacqueline He, Akari Asai et al.

Scaling laws with respect to the amount of training data and the number of parameters allow us to predict the cost-benefit trade-offs of pretraining language models (LMs) in different configurations. In this paper, we consider another dimension of scaling: the amount of data available at inference time. Specifically, we find that increasing the size of the datastore used by a retrieval-based LM monotonically improves language modeling and several downstream tasks without obvious saturation, such that a smaller model augmented with a large datastore outperforms a larger LM-only model on knowledge-intensive tasks. By plotting compute-optimal scaling curves with varied datastore, model, and pretraining data sizes, we show that using larger datastores can significantly improve model performance for the same training compute budget. We carry out our study by constructing a 1.4 trillion-token datastore named MassiveDS, which is the largest and the most diverse open-sourced datastore for retrieval-based LMs to date, and designing an efficient pipeline for studying datastore scaling in a computationally accessible manner. Finally, we analyze the effect of improving the retriever, datastore quality filtering, and other design choices on our observed scaling trends. Overall, our results show that datastore size should be considered as an integral part of LM efficiency and performance trade-offs. To facilitate future research, we open-source our datastore and code at https://github.com/RulinShao/retrieval-scaling.

Xinyan Velocity Yu, Akari Asai, Trina Chatterjee et al. · uw

While the NLP community is generally aware of resource disparities among languages, we lack research that quantifies the extent and types of such disparity. Prior surveys estimating the availability of resources based on the number of datasets can be misleading as dataset quality varies: many datasets are automatically induced or translated from English data. To provide a more comprehensive picture of language resources, we examine the characteristics of 156 publicly available NLP datasets. We manually annotate how they are created, including input text and label sources and tools used to build them, and what they study, tasks they address and motivations for their creation. After quantifying the qualitative NLP resource gap across languages, we discuss how to improve data collection in low-resource languages. We survey language-proficient NLP researchers and crowd workers per language, finding that their estimated availability correlates with dataset availability. Through crowdsourcing experiments, we identify strategies for collecting high-quality multilingual data on the Mechanical Turk platform. We conclude by making macro and micro-level suggestions to the NLP community and individual researchers for future multilingual data development.

Akari Asai, Shayne Longpre, Jungo Kasai et al. · uw

We present the results of the Workshop on Multilingual Information Access (MIA) 2022 Shared Task, evaluating cross-lingual open-retrieval question answering (QA) systems in 16 typologically diverse languages. In this task, we adapted two large-scale cross-lingual open-retrieval QA datasets in 14 typologically diverse languages, and newly annotated open-retrieval QA data in 2 underrepresented languages: Tagalog and Tamil. Four teams submitted their systems. The best system leveraging iteratively mined diverse negative examples and larger pretrained models achieves 32.2 F1, outperforming our baseline by 4.5 points. The second best system uses entity-aware contextualized representations for document retrieval, and achieves significant improvements in Tamil (20.8 F1), whereas most of the other systems yield nearly zero scores.

Tong Chen, Akari Asai, Niloofar Mireshghallah et al.

Evaluating the degree of reproduction of copyright-protected content by language models (LMs) is of significant interest to the AI and legal communities. Although both literal and non-literal similarities are considered by courts when assessing the degree of reproduction, prior research has focused only on literal similarities. To bridge this gap, we introduce CopyBench, a benchmark designed to measure both literal and non-literal copying in LM generations. Using copyrighted fiction books as text sources, we provide automatic evaluation protocols to assess literal and non-literal copying, balanced against the model utility in terms of the ability to recall facts from the copyrighted works and generate fluent completions. We find that, although literal copying is relatively rare, two types of non-literal copying -- event copying and character copying -- occur even in models as small as 7B parameters. Larger models demonstrate significantly more copying, with literal copying rates increasing from 0.2\% to 10.5\% and non-literal copying from 2.3\% to 5.9\% when comparing Llama3-8B and 70B models, respectively. We further evaluate the effectiveness of current strategies for mitigating copying and show that (1) training-time alignment can reduce literal copying but may increase non-literal copying, and (2) current inference-time mitigation methods primarily reduce literal but not non-literal copying.

Zijian Chen, Xueguang Ma, Shengyao Zhuang et al.

Deep Research agents are rapidly emerging as primary consumers of modern retrieval systems. Unlike human users who issue and refine queries without documenting their intermediate thought processes, Deep Research agents generate explicit natural language reasoning before each search call, revealing rich intent and contextual information that existing retrievers entirely ignore. To exploit this overlooked signal, we introduce: (1) Reasoning-Aware Retrieval, a retrieval paradigm that jointly embeds the agent's reasoning trace alongside its query; and (2) DR-Synth, a data synthesis method that generates Deep Research retriever training data from standard QA datasets. We demonstrate that both components are independently effective, and their combination yields a trained embedding model, AgentIR-4B, with substantial gains. On the challenging BrowseComp-Plus benchmark, AgentIR-4B achieves 68\% accuracy with the open-weight agent Tongyi-DeepResearch, compared to 50\% with conventional embedding models twice its size, and 37\% with BM25. Code and data are available at: https://texttron.github.io/AgentIR/.

Team Olmo, Allyson Ettinger, Amanda Bertsch et al. · uw

We introduce Olmo 3, a family of state-of-the-art, fully-open language models at the 7B and 32B parameter scales. Olmo 3 model construction targets long-context reasoning, function calling, coding, instruction following, general chat, and knowledge recall. This release includes the entire model flow, i.e., the full lifecycle of the family of models, including every stage, checkpoint, data point, and dependency used to build it. Our flagship model, Olmo 3 Think 32B, is the strongest fully-open thinking model released to-date.

Akari Asai, Jacqueline He, Rulin Shao et al. · allen-ai

Scientific progress depends on researchers' ability to synthesize the growing body of literature. Can large language models (LMs) assist scientists in this task? We introduce OpenScholar, a specialized retrieval-augmented LM that answers scientific queries by identifying relevant passages from 45 million open-access papers and synthesizing citation-backed responses. To evaluate OpenScholar, we develop ScholarQABench, the first large-scale multi-domain benchmark for literature search, comprising 2,967 expert-written queries and 208 long-form answers across computer science, physics, neuroscience, and biomedicine. On ScholarQABench, OpenScholar-8B outperforms GPT-4o by 5% and PaperQA2 by 7% in correctness, despite being a smaller, open model. While GPT4o hallucinates citations 78 to 90% of the time, OpenScholar achieves citation accuracy on par with human experts. OpenScholar's datastore, retriever, and self-feedback inference loop also improves off-the-shelf LMs: for instance, OpenScholar-GPT4o improves GPT-4o's correctness by 12%. In human evaluations, experts preferred OpenScholar-8B and OpenScholar-GPT4o responses over expert-written ones 51% and 70% of the time, respectively, compared to GPT4o's 32%. We open-source all of our code, models, datastore, data and a public demo.

Xiang Yue, Yueqi Song, Akari Asai et al. · cmu

Despite recent advances in multimodal large language models (MLLMs), their development has predominantly focused on English- and western-centric datasets and tasks, leaving most of the world's languages and diverse cultural contexts underrepresented. This paper introduces Pangea, a multilingual multimodal LLM trained on PangeaIns, a diverse 6M instruction dataset spanning 39 languages. PangeaIns features: 1) high-quality English instructions, 2) carefully machine-translated instructions, and 3) culturally relevant multimodal tasks to ensure cross-cultural coverage. To rigorously assess models' capabilities, we introduce PangeaBench, a holistic evaluation suite encompassing 14 datasets covering 47 languages. Results show that Pangea significantly outperforms existing open-source models in multilingual settings and diverse cultural contexts. Ablation studies further reveal the importance of English data proportions, language popularity, and the number of multimodal training samples on overall performance. We fully open-source our data, code, and trained checkpoints, to facilitate the development of inclusive and robust multilingual MLLMs, promoting equity and accessibility across a broader linguistic and cultural spectrum.

Seungone Kim, Dongkeun Yoon, Kiril Gashteovski et al.

With the advancement of AI capabilities, AI reviewers are beginning to be deployed in scientific peer review, yet their capability and credibility remain in question: many scientists simply view them as probabilistic systems without the expertise to evaluate research, while other researchers are more optimistic about their readiness without concrete evidence. Understanding what AI reviewers do well, where they fall short, and what challenges remain is essential. However, existing evaluations of AI reviewers have focused on whether their verdicts match human verdicts (e.g., score alignment, acceptance prediction), which is insufficient to characterize their capabilities and limits. In this paper, we close this gap through a large-scale expert annotation study, in which 45 domain scientists in Physical, Biological, and Health Sciences spent 469 hours rating 2,960 individual criticisms (each targeting one specific aspect of a paper) from human-written and AI-generated reviews of 82 Nature-family papers on correctness, significance, and sufficiency of evidence. On a composite of all three dimensions, a reviewing agent powered by GPT-5.2 scores above each paper's top-rated human reviewer (60.0% vs. 48.2%, p = 0.009), while all three AI reviewers (including Gemini 3.0 Pro and Claude Opus 4.5) exceed the lowest-rated human across every dimension. AI reviewers' accurate criticisms are also more often rated significant and well-evidenced, and surface a distinct 26% of issues no human raises. However, AI reviewers overlap far more than humans do (21% vs. 3% for cross-reviewer pairs), and exhibit 16 recurring weaknesses humans do not share, such as limited subfield knowledge, lack of long context management over multiple files, and overly critical stance on minor issues. Overall, our results position current AI reviewers as complements to, not substitutes for, human reviewers.

Abhika Mishra, Akari Asai, Vidhisha Balachandran et al. · cmu, uw

Large language models (LMs) are prone to generate factual errors, which are often called hallucinations. In this paper, we introduce a comprehensive taxonomy of hallucinations and argue that hallucinations manifest in diverse forms, each requiring varying degrees of careful assessments to verify factuality. We propose a novel task of automatic fine-grained hallucination detection and construct a new evaluation benchmark, FavaBench, that includes about one thousand fine-grained human judgments on three LM outputs across various domains. Our analysis reveals that ChatGPT and Llama2-Chat (70B, 7B) exhibit diverse types of hallucinations in the majority of their outputs in information-seeking scenarios. We train FAVA, a retrieval-augmented LM by carefully creating synthetic data to detect and correct fine-grained hallucinations. On our benchmark, our automatic and human evaluations show that FAVA significantly outperforms ChatGPT and GPT-4 on fine-grained hallucination detection, and edits suggested by FAVA improve the factuality of LM-generated text.

Ikuya Yamada, Akari Asai, Hannaneh Hajishirzi

Most state-of-the-art open-domain question answering systems use a neural retrieval model to encode passages into continuous vectors and extract them from a knowledge source. However, such retrieval models often require large memory to run because of the massive size of their passage index. In this paper, we introduce Binary Passage Retriever (BPR), a memory-efficient neural retrieval model that integrates a learning-to-hash technique into the state-of-the-art Dense Passage Retriever (DPR) to represent the passage index using compact binary codes rather than continuous vectors. BPR is trained with a multi-task objective over two tasks: efficient candidate generation based on binary codes and accurate reranking based on continuous vectors. Compared with DPR, BPR substantially reduces the memory cost from 65GB to 2GB without a loss of accuracy on two standard open-domain question answering benchmarks: Natural Questions and TriviaQA. Our code and trained models are available at https://github.com/studio-ousia/bpr.

Ikuya Yamada, Akari Asai, Hiroyuki Shindo et al.

Entity representations are useful in natural language tasks involving entities. In this paper, we propose new pretrained contextualized representations of words and entities based on the bidirectional transformer. The proposed model treats words and entities in a given text as independent tokens, and outputs contextualized representations of them. Our model is trained using a new pretraining task based on the masked language model of BERT. The task involves predicting randomly masked words and entities in a large entity-annotated corpus retrieved from Wikipedia. We also propose an entity-aware self-attention mechanism that is an extension of the self-attention mechanism of the transformer, and considers the types of tokens (words or entities) when computing attention scores. The proposed model achieves impressive empirical performance on a wide range of entity-related tasks. In particular, it obtains state-of-the-art results on five well-known datasets: Open Entity (entity typing), TACRED (relation classification), CoNLL-2003 (named entity recognition), ReCoRD (cloze-style question answering), and SQuAD 1.1 (extractive question answering). Our source code and pretrained representations are available at https://github.com/studio-ousia/luke.

Ikuya Yamada, Akari Asai, Jin Sakuma et al.

The embeddings of entities in a large knowledge base (e.g., Wikipedia) are highly beneficial for solving various natural language tasks that involve real world knowledge. In this paper, we present Wikipedia2Vec, a Python-based open-source tool for learning the embeddings of words and entities from Wikipedia. The proposed tool enables users to learn the embeddings efficiently by issuing a single command with a Wikipedia dump file as an argument. We also introduce a web-based demonstration of our tool that allows users to visualize and explore the learned embeddings. In our experiments, our tool achieved a state-of-the-art result on the KORE entity relatedness dataset, and competitive results on various standard benchmark datasets. Furthermore, our tool has been used as a key component in various recent studies. We publicize the source code, demonstration, and the pretrained embeddings for 12 languages at https://wikipedia2vec.github.io.

Guijin Son, Seungone Kim, Catherine Arnett et al.

Following the recent achievement of gold-medal performance on the IMO by frontier LLMs, the community is searching for the next meaningful and challenging target for measuring LLM reasoning. Whereas olympiad-style problems measure step-by-step reasoning alone, research-level problems use such reasoning to advance the frontier of mathematical knowledge itself, emerging as a compelling alternative. Yet research-level math benchmarks remain scarce because such problems are difficult to source (e.g., Riemann Bench and FrontierMath-Tier 4 contain 25 and 50 problems, respectively). To support reliable evaluation of next-generation frontier models, we introduce Soohak, a 439-problem benchmark newly authored from scratch by 64 mathematicians. Soohak comprises two subsets. On the Challenge subset, frontier models including Gemini-3-Pro, GPT-5, and Claude-Opus-4.5 reach 30.4%, 26.4%, and 10.4% respectively, leaving substantial headroom, while leading open-weight models such as Qwen3-235B, GPT-OSS-120B, and Kimi-2.5 remain below 15%. Notably, beyond standard problem solving, Soohak introduces a refusal subset that probes a capability intrinsic to research mathematics: recognizing ill-posed problems and pausing rather than producing confident but unjustified answers. On this subset, no model exceeds 50%, identifying refusal as a new optimization target that current models do not directly address. To prevent contamination, the dataset will be publicly released in late 2026, with model evaluations available upon request in the interim.

Akari Asai, Zexuan Zhong, Danqi Chen et al. · princeton, uw

Parametric language models (LMs), which are trained on vast amounts of web data, exhibit remarkable flexibility and capability. However, they still face practical challenges such as hallucinations, difficulty in adapting to new data distributions, and a lack of verifiability. In this position paper, we advocate for retrieval-augmented LMs to replace parametric LMs as the next generation of LMs. By incorporating large-scale datastores during inference, retrieval-augmented LMs can be more reliable, adaptable, and attributable. Despite their potential, retrieval-augmented LMs have yet to be widely adopted due to several obstacles: specifically, current retrieval-augmented LMs struggle to leverage helpful text beyond knowledge-intensive tasks such as question answering, have limited interaction between retrieval and LM components, and lack the infrastructure for scaling. To address these, we propose a roadmap for developing general-purpose retrieval-augmented LMs. This involves a reconsideration of datastores and retrievers, the exploration of pipelines with improved retriever-LM interaction, and significant investment in infrastructure for efficient training and inference.

Prahaladh Chandrahasan, Jiahe Jin, Zhihan Zhang et al.

Effectively evaluating deep research agents that autonomously search the web, analyze information, and generate reports remains a major challenge, particularly when it comes to assessing long reports and giving detailed feedback on their intermediate steps. To address these gaps, we introduce Deep Research Comparator, a platform that offers a holistic framework for deep research agent hosting, side-by-side comparison, fine-grained human feedback collection, and ranking calculation. Given a user query, our platform displays the final reports from two different agents along with their intermediate steps during generation. Annotators can evaluate the overall quality of final reports based on side-by-side comparison, and also provide detailed feedback separately by assessing intermediate steps or specific text spans within the final report. Furthermore, we develop Simple Deepresearch, an end-to-end agent scaffold. This scaffold serves as a baseline that facilitates the easy integration of various large language models to transform them into deep research agents for evaluation. To demonstrate the platform's utility for deep research agent development, we have collected real user preference data from 17 annotators on three deep research agents. A demo video of our platform can be found at https://www.youtube.com/watch?v=g4d2dnbdseg.

Tianyu Cao, Neel Bhandari, Akhila Yerukola et al. · allen-ai, cmu

Despite the impressive performance of Retrieval-augmented Generation (RAG) systems across various NLP benchmarks, their robustness in handling real-world user-LLM interaction queries remains largely underexplored. This presents a critical gap for practical deployment, where user queries exhibit greater linguistic variations and can trigger cascading errors across interdependent RAG components. In this work, we systematically analyze how varying four linguistic dimensions (formality, readability, politeness, and grammatical correctness) impact RAG performance. We evaluate two retrieval models and nine LLMs, ranging from 3 to 72 billion parameters, across four information-seeking Question Answering (QA) datasets. Our results reveal that linguistic reformulations significantly impact both retrieval and generation stages, leading to a relative performance drop of up to 40.41% in Recall@5 scores for less formal queries and 38.86% in answer match scores for queries containing grammatical errors. Notably, RAG systems exhibit greater sensitivity to such variations compared to LLM-only generations, highlighting their vulnerability to error propagation due to linguistic shifts. These findings highlight the need for improved robustness techniques to enhance reliability in diverse user interactions.

Rulin Shao, Akari Asai, Shannon Zejiang Shen et al.

Deep research models perform multi-step research to produce long-form, well-attributed answers. However, most open deep research models are trained on easily verifiable short-form QA tasks via reinforcement learning with verifiable rewards (RLVR), which does not extend to realistic long-form tasks. We address this with Reinforcement Learning with Evolving Rubrics (RLER), in which we construct and maintain rubrics that co-evolve with the policy model during training; this allows the rubrics to incorporate information that the model has newly explored and to provide discriminative, on-policy feedback. Using RLER, we develop Deep Research Tulu (DR Tulu-8B), the first open model that is directly trained for open-ended, long-form deep research. Across four long-form deep research benchmarks in science, healthcare and general domains, DR Tulu substantially outperforms existing open deep research models, and matches or exceeds proprietary deep research systems, while being significantly smaller and cheaper per query. To facilitate future research, we release all data, models, and code, including our new MCP-based agent infrastructure for deep research systems.

Tong Chen, Akari Asai, Luke Zettlemoyer et al.

Language models often generate factually incorrect information unsupported by their training data, a phenomenon known as extrinsic hallucination. Existing mitigation approaches often degrade performance on open-ended generation and downstream tasks, limiting their practical utility. We propose an online reinforcement learning method using a novel binary retrieval-augmented reward (RAR) to address this tradeoff. Unlike continuous reward schemes, our approach assigns a reward of one only when the model's output is entirely factually correct, and zero otherwise. We evaluate our method on Qwen3 reasoning models across diverse tasks. For open-ended generation, binary RAR achieves a 39.3% reduction in hallucination rates, substantially outperforming both supervised training and continuous-reward RL baselines. In short-form question answering, the model learns calibrated abstention, strategically outputting "I don't know" when faced with insufficient parametric knowledge. This yields 44.4% and 21.7% fewer incorrect answers on PopQA and GPQA, respectively. Crucially, these factuality gains come without performance degradation on instruction following, math, or code, whereas continuous-reward RL, despite improving factuality, induces quality regressions.

Zora Zhiruo Wang, Akari Asai, Xinyan Velocity Yu et al.

While language models (LMs) have proven remarkably adept at generating code, many programs are challenging for LMs to generate using their parametric knowledge alone. Providing external contexts such as library documentation can facilitate generating accurate and functional code. Despite the success of retrieval-augmented generation (RAG) in various text-oriented tasks, its potential for improving code generation remains under-explored. In this work, we conduct a systematic, large-scale analysis by asking: in what scenarios can retrieval benefit code generation models? and what challenges remain? We first curate a comprehensive evaluation benchmark, CodeRAG-Bench, encompassing three categories of code generation tasks, including basic programming, open-domain, and repository-level problems. We aggregate documents from five sources for models to retrieve contexts: competition solutions, online tutorials, library documentation, StackOverflow posts, and GitHub repositories. We examine top-performing models on CodeRAG-Bench by providing contexts retrieved from one or multiple sources. While notable gains are made in final code generation by retrieving high-quality contexts across various settings, our analysis reveals room for improvement -- current retrievers still struggle to fetch useful contexts especially with limited lexical overlap, and generators fail to improve with limited context lengths or abilities to integrate additional contexts. We hope CodeRAG-Bench serves as an effective testbed to encourage further development of advanced code-oriented RAG methods.

Akari Asai, Sneha Kudugunta, Xinyan Velocity Yu et al.

Despite remarkable advancements in few-shot generalization in natural language processing, most models are developed and evaluated primarily in English. To facilitate research on few-shot cross-lingual transfer, we introduce a new benchmark, called BUFFET, which unifies 15 diverse tasks across 54 languages in a sequence-to-sequence format and provides a fixed set of few-shot examples and instructions. BUFFET is designed to establish a rigorous and equitable evaluation framework for few-shot cross-lingual transfer across a broad range of tasks and languages. Using BUFFET, we perform thorough evaluations of state-of-the-art multilingual large language models with different transfer methods, namely in-context learning and fine-tuning. Our findings reveal significant room for improvement in few-shot in-context cross-lingual transfer. In particular, ChatGPT with in-context learning often performs worse than much smaller mT5-base models fine-tuned on English task data and few-shot in-language examples. Our analysis suggests various avenues for future research in few-shot cross-lingual transfer, such as improved pretraining, understanding, and future evaluations.

Joongwon Kim, Akari Asai, Gabriel Ilharco et al.

Recent work in NLP has shown promising results in training models on large amounts of tasks to achieve better generalization. However, it is not well-understood how tasks are related, and how helpful training tasks can be chosen for a new task. In this work, we investigate whether knowing task relationships via pairwise task transfer improves choosing one or more source tasks that help to learn a new target task. We provide TaskWeb, a large-scale benchmark of pairwise task transfers for 22 NLP tasks using three different model types, sizes, and adaptation methods, spanning about 25,000 experiments. Then, we design a new method TaskShop based on our analysis of TaskWeb. TaskShop uses TaskWeb to estimate the benefit of using a source task for learning a new target task, and to choose a subset of helpful training tasks for multi-task training. Our method improves overall rankings and top-k precision of source tasks by 10% and 38%, respectively. We also use TaskShop to build much smaller multi-task training sets that improve zero-shot performances across 11 different target tasks by at least 4.3%.

Xiaoyu Shen, Akari Asai, Bill Byrne et al.

Product Question Answering (PQA) systems are key in e-commerce applications to provide responses to customers' questions as they shop for products. While existing work on PQA focuses mainly on English, in practice there is need to support multiple customer languages while leveraging product information available in English. To study this practical industrial task, we present xPQA, a large-scale annotated cross-lingual PQA dataset in 12 languages across 9 branches, and report results in (1) candidate ranking, to select the best English candidate containing the information to answer a non-English question; and (2) answer generation, to generate a natural-sounding non-English answer based on the selected English candidate. We evaluate various approaches involving machine translation at runtime or offline, leveraging multilingual pre-trained LMs, and including or excluding xPQA training data. We find that (1) In-domain data is essential as cross-lingual rankers trained on other domains perform poorly on the PQA task; (2) Candidate ranking often prefers runtime-translation approaches while answer generation prefers multilingual approaches; (3) Translating offline to augment multilingual models helps candidate ranking mainly on languages with non-Latin scripts; and helps answer generation mainly on languages with Latin scripts. Still, there remains a significant performance gap between the English and the cross-lingual test sets.

Odunayo Ogundepo, Tajuddeen R. Gwadabe, Clara E. Rivera et al.

African languages have far less in-language content available digitally, making it challenging for question answering systems to satisfy the information needs of users. Cross-lingual open-retrieval question answering (XOR QA) systems -- those that retrieve answer content from other languages while serving people in their native language -- offer a means of filling this gap. To this end, we create AfriQA, the first cross-lingual QA dataset with a focus on African languages. AfriQA includes 12,000+ XOR QA examples across 10 African languages. While previous datasets have focused primarily on languages where cross-lingual QA augments coverage from the target language, AfriQA focuses on languages where cross-lingual answer content is the only high-coverage source of answer content. Because of this, we argue that African languages are one of the most important and realistic use cases for XOR QA. Our experiments demonstrate the poor performance of automatic translation and multilingual retrieval methods. Overall, AfriQA proves challenging for state-of-the-art QA models. We hope that the dataset enables the development of more equitable QA technology.

Akari Asai, Matt Gardner, Hannaneh Hajishirzi

Retrieval-augmented generation models have shown state-of-the-art performance across many knowledge-intensive NLP tasks such as open question answering and fact verification. These models are trained to generate the final output given the retrieved passages, which can be irrelevant to the original query, leading to learning spurious cues or answer memorization. This work introduces a method to incorporate the evidentiality of passages -- whether a passage contains correct evidence to support the output -- into training the generator. We introduce a multi-task learning framework to jointly generate the final output and predict the evidentiality of each passage, leveraging a new task-agnostic method to obtain silver evidentiality labels for supervision. Our experiments on five datasets across three knowledge-intensive tasks show that our new evidentiality-guided generator significantly outperforms its direct counterpart with the same-size model and advances the state of the art on FaVIQ-Ambig. We attribute these improvements to both the auxiliary multi-task learning and silver evidentiality mining techniques.

Akari Asai, Xinyan Yu, Jungo Kasai et al.

We present Cross-lingual Open-Retrieval Answer Generation (CORA), the first unified many-to-many question answering (QA) model that can answer questions across many languages, even for ones without language-specific annotated data or knowledge sources. We introduce a new dense passage retrieval algorithm that is trained to retrieve documents across languages for a question. Combined with a multilingual autoregressive generation model, CORA answers directly in the target language without any translation or in-language retrieval modules as used in prior work. We propose an iterative training method that automatically extends annotated data available only in high-resource languages to low-resource ones. Our results show that CORA substantially outperforms the previous state of the art on multilingual open QA benchmarks across 26 languages, 9 of which are unseen during training. Our analyses show the significance of cross-lingual retrieval and generation in many languages, particularly under low-resource settings.

Alon Talmor, Ori Yoran, Amnon Catav et al.

When answering complex questions, people can seamlessly combine information from visual, textual and tabular sources. While interest in models that reason over multiple pieces of evidence has surged in recent years, there has been relatively little work on question answering models that reason across multiple modalities. In this paper, we present MultiModalQA(MMQA): a challenging question answering dataset that requires joint reasoning over text, tables and images. We create MMQA using a new framework for generating complex multi-modal questions at scale, harvesting tables from Wikipedia, and attaching images and text paragraphs using entities that appear in each table. We then define a formal language that allows us to take questions that can be answered from a single modality, and combine them to generate cross-modal questions. Last, crowdsourcing workers take these automatically-generated questions and rephrase them into more fluent language. We create 29,918 questions through this procedure, and empirically demonstrate the necessity of a multi-modal multi-hop approach to solve our task: our multi-hop model, ImplicitDecomp, achieves an average F1of 51.7 over cross-modal questions, substantially outperforming a strong baseline that achieves 38.2 F1, but still lags significantly behind human performance, which is at 90.1 F1

Takumi Kawashima, Qing Yu, Akari Asai et al.

We propose a new optimization framework for aleatoric uncertainty estimation in regression problems. Existing methods can quantify the error in the target estimation, but they tend to underestimate it. To obtain the predictive uncertainty inherent in an observation, we propose a new separable formulation for the estimation of a signal and of its uncertainty, avoiding the effect of overfitting. By decoupling target estimation and uncertainty estimation, we also control the balance between signal estimation and uncertainty estimation. We conduct three types of experiments: regression with simulation data, age estimation, and depth estimation. We demonstrate that the proposed method outperforms a state-of-the-art technique for signal and uncertainty estimation.

Akari Asai, Eunsol Choi

Recent pretrained language models "solved" many reading comprehension benchmarks, where questions are written with access to the evidence document. However, datasets containing information-seeking queries where evidence documents are provided after the queries are written independently remain challenging. We analyze why answering information-seeking queries is more challenging and where their prevalent unanswerabilities arise, on Natural Questions and TyDi QA. Our controlled experiments suggest two headrooms -- paragraph selection and answerability prediction, i.e. whether the paired evidence document contains the answer to the query or not. When provided with a gold paragraph and knowing when to abstain from answering, existing models easily outperform a human annotator. However, predicting answerability itself remains challenging. We manually annotate 800 unanswerable examples across six languages on what makes them challenging to answer. With this new data, we conduct per-category answerability prediction, revealing issues in the current dataset collection as well as task formulation. Together, our study points to avenues for future research in information-seeking question answering, both for dataset creation and model development.

Akari Asai, Jungo Kasai, Jonathan H. Clark et al.

Multilingual question answering tasks typically assume answers exist in the same language as the question. Yet in practice, many languages face both information scarcity -- where languages have few reference articles -- and information asymmetry -- where questions reference concepts from other cultures. This work extends open-retrieval question answering to a cross-lingual setting enabling questions from one language to be answered via answer content from another language. We construct a large-scale dataset built on questions from TyDi QA lacking same-language answers. Our task formulation, called Cross-lingual Open Retrieval Question Answering (XOR QA), includes 40k information-seeking questions from across 7 diverse non-English languages. Based on this dataset, we introduce three new tasks that involve cross-lingual document retrieval using multi-lingual and English resources. We establish baselines with state-of-the-art machine translation systems and cross-lingual pretrained models. Experimental results suggest that XOR QA is a challenging task that will facilitate the development of novel techniques for multilingual question answering. Our data and code are available at https://nlp.cs.washington.edu/xorqa.

Akari Asai, Hannaneh Hajishirzi

Many natural language questions require qualitative, quantitative or logical comparisons between two entities or events. This paper addresses the problem of improving the accuracy and consistency of responses to comparison questions by integrating logic rules and neural models. Our method leverages logical and linguistic knowledge to augment labeled training data and then uses a consistency-based regularizer to train the model. Improving the global consistency of predictions, our approach achieves large improvements over previous methods in a variety of question answering (QA) tasks including multiple-choice qualitative reasoning, cause-effect reasoning, and extractive machine reading comprehension. In particular, our method significantly improves the performance of RoBERTa-based models by 1-5% across datasets. We advance the state of the art by around 5-8% on WIQA and QuaRel and reduce consistency violations by 58% on HotpotQA. We further demonstrate that our approach can learn effectively from limited data.

Daya Guo, Akari Asai, Duyu Tang et al.

We study the problem of generating inferential texts of events for a variety of commonsense like \textit{if-else} relations. Existing approaches typically use limited evidence from training examples and learn for each relation individually. In this work, we use multiple knowledge sources as fuels for the model. Existing commonsense knowledge bases like ConceptNet are dominated by taxonomic knowledge (e.g., \textit{isA} and \textit{relatedTo} relations), having a limited number of inferential knowledge. We use not only structured commonsense knowledge bases, but also natural language snippets from search-engine results. These sources are incorporated into a generative base model via key-value memory network. In addition, we introduce a meta-learning based multi-task learning algorithm. For each targeted commonsense relation, we regard the learning of examples from other relations as the meta-training process, and the evaluation on examples from the targeted relation as the meta-test process. We conduct experiments on Event2Mind and ATOMIC datasets. Results show that both the integration of multiple knowledge sources and the use of the meta-learning algorithm improve the performance.

Lichao Sun, Kazuma Hashimoto, Wenpeng Yin et al.

There is an increasing amount of literature that claims the brittleness of deep neural networks in dealing with adversarial examples that are created maliciously. It is unclear, however, how the models will perform in realistic scenarios where \textit{natural rather than malicious} adversarial instances often exist. This work systematically explores the robustness of BERT, the state-of-the-art Transformer-style model in NLP, in dealing with noisy data, particularly mistakes in typing the keyboard, that occur inadvertently. Intensive experiments on sentiment analysis and question answering benchmarks indicate that: (i) Typos in various words of a sentence do not influence equally. The typos in informative words make severer damages; (ii) Mistype is the most damaging factor, compared with inserting, deleting, etc.; (iii) Humans and machines have different focuses on recognizing adversarial attacks.

Akari Asai, Kazuma Hashimoto, Hannaneh Hajishirzi et al.

Answering questions that require multi-hop reasoning at web-scale necessitates retrieving multiple evidence documents, one of which often has little lexical or semantic relationship to the question. This paper introduces a new graph-based recurrent retrieval approach that learns to retrieve reasoning paths over the Wikipedia graph to answer multi-hop open-domain questions. Our retriever model trains a recurrent neural network that learns to sequentially retrieve evidence paragraphs in the reasoning path by conditioning on the previously retrieved documents. Our reader model ranks the reasoning paths and extracts the answer span included in the best reasoning path. Experimental results show state-of-the-art results in three open-domain QA datasets, showcasing the effectiveness and robustness of our method. Notably, our method achieves significant improvement in HotpotQA, outperforming the previous best model by more than 14 points.

Akari Asai, Akiko Eriguchi, Kazuma Hashimoto et al.

Despite recent work in Reading Comprehension (RC), progress has been mostly limited to English due to the lack of large-scale datasets in other languages. In this work, we introduce the first RC system for languages without RC training data. Given a target language without RC training data and a pivot language with RC training data (e.g. English), our method leverages existing RC resources in the pivot language by combining a competitive RC model in the pivot language with an attentive Neural Machine Translation (NMT) model. We first translate the data from the target to the pivot language, and then obtain an answer using the RC model in the pivot language. Finally, we recover the corresponding answer in the original language using soft-alignment attention scores from the NMT model. We create evaluation sets of RC data in two non-English languages, namely Japanese and French, to evaluate our method. Experimental results on these datasets show that our method significantly outperforms a back-translation baseline of a state-of-the-art product-level machine translation system.

Akari Asai, Sara Evensen, Behzad Golshan et al.

The science of happiness is an area of positive psychology concerned with understanding what behaviors make people happy in a sustainable fashion. Recently, there has been interest in developing technologies that help incorporate the findings of the science of happiness into users' daily lives by steering them towards behaviors that increase happiness. With the goal of building technology that can understand how people express their happy moments in text, we crowd-sourced HappyDB, a corpus of 100,000 happy moments that we make publicly available. This paper describes HappyDB and its properties, and outlines several important NLP problems that can be studied with the help of the corpus. We also apply several state-of-the-art analysis techniques to analyze HappyDB. Our results demonstrate the need for deeper NLP techniques to be developed which makes HappyDB an exciting resource for follow-on research.