Doug Downey

Rodney Kinney, Chloe Anastasiades, Russell Authur et al. · allen-ai, microsoft-research

The volume of scientific output is creating an urgent need for automated tools to help scientists keep up with developments in their field. Semantic Scholar (S2) is an open data platform and website aimed at accelerating science by helping scholars discover and understand scientific literature. We combine public and proprietary data sources using state-of-the-art techniques for scholarly PDF content extraction and automatic knowledge graph construction to build the Semantic Scholar Academic Graph, the largest open scientific literature graph to-date, with 200M+ papers, 80M+ authors, 550M+ paper-authorship edges, and 2.4B+ citation edges. The graph includes advanced semantic features such as structurally parsed text, natural language summaries, and vector embeddings. In this paper, we describe the components of the S2 data processing pipeline and the associated APIs offered by the platform. We will update this living document to reflect changes as we add new data offerings and improve existing services.

Daniel King, Zejiang Shen, Nishant Subramani et al. · allen-ai, cmu

Abstractive summarization systems today produce fluent and relevant output, but often "hallucinate" statements not supported by the source text. We analyze the connection between hallucinations and training data, and find evidence that models hallucinate because they train on target summaries that are unsupported by the source. Based on our findings, we present PINOCCHIO, a new decoding method that improves the consistency of a transformer-based abstractive summarizer by constraining beam search to avoid hallucinations. Given the model states and outputs at a given step, PINOCCHIO detects likely model hallucinations based on various measures of attribution to the source text. PINOCCHIO backtracks to find more consistent output, and can opt to produce no summary at all when no consistent generation can be found. In experiments, we find that PINOCCHIO improves the consistency of generation (in terms of F1) by an average of~67% on two abstractive summarization datasets.

Kyle Lo, Joseph Chee Chang, Andrew Head et al. · allen-ai, cmu

Scholarly publications are key to the transfer of knowledge from scholars to others. However, research papers are information-dense, and as the volume of the scientific literature grows, the need for new technology to support the reading process grows. In contrast to the process of finding papers, which has been transformed by Internet technology, the experience of reading research papers has changed little in decades. The PDF format for sharing research papers is widely used due to its portability, but it has significant downsides including: static content, poor accessibility for low-vision readers, and difficulty reading on mobile devices. This paper explores the question "Can recent advances in AI and HCI power intelligent, interactive, and accessible reading interfaces -- even for legacy PDFs?" We describe the Semantic Reader Project, a collaborative effort across multiple institutions to explore automatic creation of dynamic reading interfaces for research papers. Through this project, we've developed ten research prototype interfaces and conducted usability studies with more than 300 participants and real-world users showing improved reading experiences for scholars. We've also released a production reading interface for research papers that will incorporate the best features as they mature. We structure this paper around challenges scholars and the public face when reading research papers -- Discovery, Efficiency, Comprehension, Synthesis, and Accessibility -- and present an overview of our progress and remaining open challenges.

Tom Hope, Doug Downey, Oren Etzioni et al. · uw

We stand at the foot of a significant inflection in the trajectory of scientific discovery. As society continues on its fast-paced digital transformation, so does humankind's collective scientific knowledge and discourse. We now read and write papers in digitized form, and a great deal of the formal and informal processes of science are captured digitally -- including papers, preprints and books, code and datasets, conference presentations, and interactions in social networks and collaboration and communication platforms. The transition has led to the creation and growth of a tremendous amount of information -- much of which is available for public access -- opening exciting opportunities for computational models and systems that analyze and harness it. In parallel, exponential growth in data processing power has fueled remarkable advances in artificial intelligence, including large neural language models capable of learning powerful representations from unstructured text. Dramatic changes in scientific communication -- such as the advent of the first scientific journal in the 17th century -- have historically catalyzed revolutions in scientific thought. The confluence of societal and computational trends suggests that computer science is poised to ignite a revolution in the scientific process itself.

Zejiang Shen, Kyle Lo, Lauren Yu et al. · mit

With the advent of large language models, methods for abstractive summarization have made great strides, creating potential for use in applications to aid knowledge workers processing unwieldy document collections. One such setting is the Civil Rights Litigation Clearinghouse (CRLC) (https://clearinghouse.net),which posts information about large-scale civil rights lawsuits, serving lawyers, scholars, and the general public. Today, summarization in the CRLC requires extensive training of lawyers and law students who spend hours per case understanding multiple relevant documents in order to produce high-quality summaries of key events and outcomes. Motivated by this ongoing real-world summarization effort, we introduce Multi-LexSum, a collection of 9,280 expert-authored summaries drawn from ongoing CRLC writing. Multi-LexSum presents a challenging multi-document summarization task given the length of the source documents, often exceeding two hundred pages per case. Furthermore, Multi-LexSum is distinct from other datasets in its multiple target summaries, each at a different granularity (ranging from one-sentence "extreme" summaries to multi-paragraph narrations of over five hundred words). We present extensive analysis demonstrating that despite the high-quality summaries in the training data (adhering to strict content and style guidelines), state-of-the-art summarization models perform poorly on this task. We release Multi-LexSum for further research in summarization methods as well as to facilitate development of applications to assist in the CRLC's mission at https://multilexsum.github.io.

Chandra Bhagavatula, Jena D. Hwang, Doug Downey et al. · allen-ai, uw

Commonsense capabilities of pre-trained language models dramatically improve with scale, leading many to believe that scale is the only winning recipe. But is it? Here, we investigate an alternative that a priori seems impossible: can smaller language models (e.g., GPT-2) win over models that are orders of magnitude larger and better (e.g., GPT-3), if powered with novel commonsense distillation algorithms? The key intellectual challenge is to design a learning algorithm that achieve a competitive level of commonsense acquisition, without relying on the benefits of scale. In particular, we study generative models of commonsense knowledge, focusing on the task of generating generics, statements of commonsense facts about everyday concepts, e.g., birds can fly. We introduce I2D2, a novel commonsense distillation framework that loosely follows the Symbolic Knowledge Distillation of West et al. but breaks the dependence on the extreme-scale teacher model with two innovations: (1) the novel adaptation of NeuroLogic Decoding to enhance the generation quality of the weak, off-the-shelf language models, and (2) self-imitation learning to iteratively learn from the model's own enhanced commonsense acquisition capabilities. Empirical results suggest that scale is not the only way, as novel algorithms can be a promising alternative. Moreover, our study leads to a new corpus of generics, Gen-A-tomic, that is the largest and highest quality available to date.

Catherine Chen, Zejiang Shen, Dan Klein et al. · berkeley, mit

Recent work has shown that infusing layout features into language models (LMs) improves processing of visually-rich documents such as scientific papers. Layout-infused LMs are often evaluated on documents with familiar layout features (e.g., papers from the same publisher), but in practice models encounter documents with unfamiliar distributions of layout features, such as new combinations of text sizes and styles, or new spatial configurations of textual elements. In this work we test whether layout-infused LMs are robust to layout distribution shifts. As a case study we use the task of scientific document structure recovery, segmenting a scientific paper into its structural categories (e.g., "title", "caption", "reference"). To emulate distribution shifts that occur in practice we re-partition the GROTOAP2 dataset. We find that under layout distribution shifts model performance degrades by up to 20 F1. Simple training strategies, such as increasing training diversity, can reduce this degradation by over 35% relative F1; however, models fail to reach in-distribution performance in any tested out-of-distribution conditions. This work highlights the need to consider layout distribution shifts during model evaluation, and presents a methodology for conducting such evaluations.

Zejiang Shen, Tal August, Pao Siangliulue et al. · allen-ai, cmu

Large language models have introduced exciting new opportunities and challenges in designing and developing new AI-assisted writing support tools. Recent work has shown that leveraging this new technology can transform writing in many scenarios such as ideation during creative writing, editing support, and summarization. However, AI-supported expository writing--including real-world tasks like scholars writing literature reviews or doctors writing progress notes--is relatively understudied. In this position paper, we argue that developing AI supports for expository writing has unique and exciting research challenges and can lead to high real-world impacts. We characterize expository writing as evidence-based and knowledge-generating: it contains summaries of external documents as well as new information or knowledge. It can be seen as the product of authors' sensemaking process over a set of source documents, and the interplay between reading, reflection, and writing opens up new opportunities for designing AI support. We sketch three components for AI support design and discuss considerations for future research.

Sonia K. Murthy, Kyle Lo, Daniel King et al. · allen-ai, uw

Systems that can automatically define unfamiliar terms hold the promise of improving the accessibility of scientific texts, especially for readers who may lack prerequisite background knowledge. However, current systems assume a single "best" description per concept, which fails to account for the many potentially useful ways a concept can be described. We present ACCoRD, an end-to-end system tackling the novel task of generating sets of descriptions of scientific concepts. Our system takes advantage of the myriad ways a concept is mentioned across the scientific literature to produce distinct, diverse descriptions of target scientific concepts in terms of different reference concepts. To support research on the task, we release an expert-annotated resource, the ACCoRD corpus, which includes 1,275 labeled contexts and 1,787 hand-authored concept descriptions. We conduct a user study demonstrating that (1) users prefer descriptions produced by our end-to-end system, and (2) users prefer multiple descriptions to a single "best" description.

Tara Safavi, Doug Downey, Tom Hope

Knowledge graph (KG) link prediction is a fundamental task in artificial intelligence, with applications in natural language processing, information retrieval, and biomedicine. Recently, promising results have been achieved by leveraging cross-modal information in KGs, using ensembles that combine knowledge graph embeddings (KGEs) and contextual language models (LMs). However, existing ensembles are either (1) not consistently effective in terms of ranking accuracy gains or (2) impractically inefficient on larger datasets due to the combinatorial explosion problem of pairwise ranking with deep language models. In this paper, we propose a novel tiered ranking architecture CascadER to maintain the ranking accuracy of full ensembling while improving efficiency considerably. CascadER uses LMs to rerank the outputs of more efficient base KGEs, relying on an adaptive subset selection scheme aimed at invoking the LMs minimally while maximizing accuracy gain over the KGE. Extensive experiments demonstrate that CascadER improves MRR by up to 9 points over KGE baselines, setting new state-of-the-art performance on four benchmarks while improving efficiency by one or more orders of magnitude over competitive cross-modal baselines. Our empirical analyses reveal that diversity of models across modalities and preservation of individual models' confidence signals help explain the effectiveness of CascadER, and suggest promising directions for cross-modal cascaded architectures. Code and pretrained models are available at https://github.com/tsafavi/cascader.

Yuze Lou, Bailey Kuehl, Erin Bransom et al. · allen-ai, cmu

Entity linking (EL) is the task of linking a textual mention to its corresponding entry in a knowledge base, and is critical for many knowledge-intensive NLP applications. When applied to tables in scientific papers, EL is a step toward large-scale scientific knowledge bases that could enable advanced scientific question answering and analytics. We present the first dataset for EL in scientific tables. EL for scientific tables is especially challenging because scientific knowledge bases can be very incomplete, and disambiguating table mentions typically requires understanding the papers's tet in addition to the table. Our dataset, S2abEL, focuses on EL in machine learning results tables and includes hand-labeled cell types, attributed sources, and entity links from the PaperswithCode taxonomy for 8,429 cells from 732 tables. We introduce a neural baseline method designed for EL on scientific tables containing many out-of-knowledge-base mentions, and show that it significantly outperforms a state-of-the-art generic table EL method. The best baselines fall below human performance, and our analysis highlights avenues for improvement.

Jon Saad-Falcon, Amanpreet Singh, Luca Soldaini et al. · allen-ai, stanford

Real-world applications of neural language models often involve running many different models over the same corpus. The high computational cost of these runs has led to interest in techniques that can reuse the contextualized embeddings produced in previous runs to speed training and inference of future ones. We refer to this approach as embedding recycling (ER). While multiple ER techniques have been proposed, their practical effectiveness is still unknown because existing evaluations consider very few models and do not adequately account for overhead costs. We perform an extensive evaluation of ER across eight different models (17 to 900 million parameters) and fourteen tasks in English. We show how a simple ER technique that caches activations from an intermediate layer of a pretrained model, and learns task-specific adapters on the later layers, is broadly effective. For the best-performing baseline in our experiments (DeBERTa-v2 XL), adding a precomputed cache results in a >90% speedup during training and 87-91% speedup for inference, with negligible impact on accuracy. Our analysis reveals important areas of future work.

Mike D'Arcy, Alexis Ross, Erin Bransom et al. · allen-ai

We introduce the task of automatically revising scientific papers based on peer feedback and release ARIES, a dataset of review comments and their corresponding paper edits. The data is drawn from real reviewer-author interactions from computer science, and we provide labels linking each reviewer comment to the specific paper edits made by the author in response. We automatically create a high-precision silver training set, as well as an expert-labeled test set that shows high inter-annotator agreement. In experiments with 10 models covering the state of the art, we find that they struggle even to identify which edits correspond to a comment -- especially when the relationship between the edit and the comment is indirect and requires reasoning to uncover. We also extensively analyze GPT-4's ability to generate edits given a comment and the original paper. We find that it often succeeds on a superficial level, but tends to rigidly follow the wording of the feedback rather than the underlying intent, and lacks technical details compared to human-written edits.

Arie Cattan, Tom Hope, Doug Downey et al.

Various NLP tasks require a complex hierarchical structure over nodes, where each node is a cluster of items. Examples include generating entailment graphs, hierarchical cross-document coreference resolution, annotating event and subevent relations, etc. To enable efficient annotation of such hierarchical structures, we release CHAMP, an open source tool allowing to incrementally construct both clusters and hierarchy simultaneously over any type of texts. This incremental approach significantly reduces annotation time compared to the common pairwise annotation approach and also guarantees maintaining transitivity at the cluster and hierarchy levels. Furthermore, CHAMP includes a consolidation mode, where an adjudicator can easily compare multiple cluster hierarchy annotations and resolve disagreements.

Emily Allaway, Jena D. Hwang, Chandra Bhagavatula et al.

Generics express generalizations about the world (e.g., birds can fly) that are not universally true (e.g., newborn birds and penguins cannot fly). Commonsense knowledge bases, used extensively in NLP, encode some generic knowledge but rarely enumerate such exceptions and knowing when a generic statement holds or does not hold true is crucial for developing a comprehensive understanding of generics. We present a novel framework informed by linguistic theory to generate exemplars -- specific cases when a generic holds true or false. We generate ~19k exemplars for ~650 generics and show that our framework outperforms a strong GPT-3 baseline by 12.8 precision points. Our analysis highlights the importance of linguistic theory-based controllability for generating exemplars, the insufficiency of knowledge bases as a source of exemplars, and the challenges exemplars pose for the task of natural language inference.

Aakanksha Naik, Bailey Kuehl, Erin Bransom et al. · allen-ai, cmu

Extracting fine-grained experimental findings from literature can provide dramatic utility for scientific applications. Prior work has developed annotation schemas and datasets for limited aspects of this problem, failing to capture the real-world complexity and nuance required. Focusing on biomedicine, this work presents CARE -- a new IE dataset for the task of extracting clinical findings. We develop a new annotation schema capturing fine-grained findings as n-ary relations between entities and attributes, which unifies phenomena challenging for current IE systems such as discontinuous entity spans, nested relations, variable arity n-ary relations and numeric results in a single schema. We collect extensive annotations for 700 abstracts from two sources: clinical trials and case reports. We also demonstrate the generalizability of our schema to the computer science and materials science domains. We benchmark state-of-the-art IE systems on CARE, showing that even models such as GPT4 struggle. We release our resources to advance research on extracting and aggregating literature findings.

Amanpreet Singh, Mike D'Arcy, Arman Cohan et al.

Learned representations of scientific documents can serve as valuable input features for downstream tasks without further fine-tuning. However, existing benchmarks for evaluating these representations fail to capture the diversity of relevant tasks. In response, we introduce SciRepEval, the first comprehensive benchmark for training and evaluating scientific document representations. It includes 24 challenging and realistic tasks, 8 of which are new, across four formats: classification, regression, ranking and search. We then use this benchmark to study and improve the generalization ability of scientific document representation models. We show how state-of-the-art models like SPECTER and SciNCL struggle to generalize across the task formats, and that simple multi-task training fails to improve them. However, a new approach that learns multiple embeddings per document, each tailored to a different format, can improve performance. We experiment with task-format-specific control codes and adapters and find they outperform the existing single-embedding state-of-the-art by over 2 points absolute. We release the resulting family of multi-format models, called SPECTER2, for the community to use and build on.

Victor S. Bursztyn, David Demeter, Doug Downey et al.

How to usefully encode compositional task structure has long been a core challenge in AI. Recent work in chain of thought prompting has shown that for very large neural language models (LMs), explicitly demonstrating the inferential steps involved in a target task may improve performance over end-to-end learning that focuses on the target task alone. However, chain of thought prompting has significant limitations due to its dependency on huge pretrained LMs. In this work, we present compositional fine-tuning (CFT): an approach based on explicitly decomposing a target task into component tasks, and then fine-tuning smaller LMs on a curriculum of such component tasks. We apply CFT to recommendation tasks in two domains, world travel and local dining, as well as a previously studied inferential task (sports understanding). We show that CFT outperforms end-to-end learning even with equal amounts of data, and gets consistently better as more component tasks are modeled via fine-tuning. Compared with chain of thought prompting, CFT performs at least as well using LMs only 7.4% of the size, and is moreover applicable to task domains for which data are not available during pretraining.

Dany Haddad, Dan Bareket, Joseph Chee Chang et al.

AI-powered scientific research tools are rapidly being integrated into research workflows, yet the field lacks a clear lens into how researchers use these systems in real-world settings. We present and analyze the Asta Interaction Dataset, a large-scale resource comprising over 200,000 user queries and interaction logs from two deployed tools (a literature discovery interface and a scientific question-answering interface) within an LLM-powered retrieval-augmented generation platform. Using this dataset, we characterize query patterns, engagement behaviors, and how usage evolves with experience. We find that users submit longer and more complex queries than in traditional search, and treat the system as a collaborative research partner, delegating tasks such as drafting content and identifying research gaps. Users treat generated responses as persistent artifacts, revisiting and navigating among outputs and cited evidence in non-linear ways. With experience, users issue more targeted queries and engage more deeply with supporting citations, although keyword-style queries persist even among experienced users. We release the anonymized dataset and analysis with a new query intent taxonomy to inform future designs of real-world AI research assistants and to support realistic evaluation.

Anirudh Ajith, Amanpreet Singh, Jay DeYoung et al.

Can AI systems trained on the scientific record up to a fixed point in time forecast the scientific advances that follow? Such a capability could help researchers identify collaborators and impactful research directions, and anticipate which problems and methods will become central next. We introduce PreScience -- a scientific forecasting benchmark that decomposes the research process into four interdependent generative tasks: collaborator prediction, prior work selection, contribution generation, and impact prediction. PreScience is a carefully curated dataset of 98K recent AI-related research papers, featuring disambiguated author identities, temporally aligned scholarly metadata, and a structured graph of companion author publication histories and citations spanning 502K total papers. We develop baselines and evaluations for each task, including LACERScore, a novel LLM-based measure of contribution similarity that outperforms previous metrics and approximates inter-annotator agreement. We find substantial headroom remains in each task -- e.g. in contribution generation, frontier LLMs achieve only moderate similarity to the ground-truth (GPT-5, averages 5.6 on a 1-10 scale). When composed into a 12-month end-to-end simulation of scientific production, the resulting synthetic corpus is systematically less diverse and less novel than human-authored research from the same period.

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.

Amanpreet Singh, Joseph Chee Chang, Chloe Anastasiades et al. · allen-ai

Retrieval-augmented generation is increasingly effective in answering scientific questions from literature, but many state-of-the-art systems are expensive and closed-source. We introduce Ai2 Scholar QA, a free online scientific question answering application. To facilitate research, we make our entire pipeline public: as a customizable open-source Python package and interactive web app, along with paper indexes accessible through public APIs and downloadable datasets. We describe our system in detail and present experiments analyzing its key design decisions. In an evaluation on a recent scientific QA benchmark, we find that Ai2 Scholar QA outperforms competing systems.

Nishant Balepur, Malachi Hamada, Varsha Kishore et al.

Scientific Deep Research (DR) agents answer user queries by synthesizing research papers into multi-section reports. User feedback can improve their utility, but existing protocols only score the final report, making it hard to study and learn which intermediate actions DR agents should take to improve reports. We collect DRACULA, the first dataset with user feedback on intermediate actions for DR. Over five weeks, nineteen expert CS researchers ask queries to a DR system that proposes actions (e.g., "Add a section on datasets"). Our users select actions they prefer, then judge whether an output report applied their selections successfully, yielding 8,103 action preferences and 5,230 execution judgments. After confirming a DR agent can execute DRACULA's actions, we study the predictability of user-preferred actions via simulation-how well LLMs predict the actions users select-a step toward learning to generate useful actions. We discover: (1) LLM judges initially struggle to predict action selections, but improve most when using a user's full selection history, rather than self-reported or extrapolated user context signals; (2) Users' selections for the same query differ based on unstated goals, bottlenecking simulation and motivating affordances that let users steer reports; and (3) Our simulation results inform an online intervention that generates new actions based on the user's past interactions, which users pick most often in follow-up studies. Overall, while work extensively studies execution, DRACULA reveals a key challenge is deciding which actions to execute in the first place. We open-source DRACULA's study design, user feedback, and simulation tasks to spur future work on action feedback for long-horizon agents.

Yilun Zhao, Kaiyan Zhang, Tiansheng Hu et al. · allen-ai

We present SciArena, an open and collaborative platform for evaluating foundation models on scientific literature tasks. Unlike traditional benchmarks for scientific literature understanding and synthesis, SciArena engages the research community directly, following the Chatbot Arena evaluation approach of community voting on model comparisons. By leveraging collective intelligence, SciArena offers a community-driven evaluation of model performance on open-ended scientific tasks that demand literature-grounded, long-form responses. The platform currently supports 23 open-source and proprietary foundation models and has collected over 13,000 votes from trusted researchers across diverse scientific domains. We analyze the data collected so far and confirm that the submitted questions are diverse, aligned with real-world literature needs, and that participating researchers demonstrate strong self-consistency and inter-annotator agreement in their evaluations. We discuss the results and insights based on the model ranking leaderboard. To further promote research in building model-based automated evaluation systems for literature tasks, we release SciArena-Eval, a meta-evaluation benchmark based on our collected preference data. The benchmark measures the accuracy of models in judging answer quality by comparing their pairwise assessments with human votes. Our experiments highlight the benchmark's challenges and emphasize the need for more reliable automated evaluation methods.

Pao Siangliulue, Jonathan Bragg, Doug Downey et al.

As AI agents become increasingly capable of complex knowledge tasks, the lack of context limits their capability to proactively reason about a user's latent needs throughout a long evolving project. In scientific research, many researchers still manually query a deep research system and compress their rich project contexts into short, targeted queries. Further, a deep research system produces exhaustive reports, making it difficult to identify concrete actions. To explore the opportunities of research assistants that are proactive throughout a research project, we conducted several studies (N=42) with a technology probe and an iterative prototype. The latest iteration of our system, Omakase, is a research assistant that monitors a user's project documents to infer timely queries to a deep research system. Omakase then distills long reports into suggestions contextualized to their evolving projects. Our evaluations showed that participants found the generated queries to be useful and timely, and rated Omakase's suggestions as significantly more actionable than the original reports.

Satanu Ghosh, Collin Holgate, Neal R. Brodnik et al.

We apply preference learning to the task of language model-guided design of novel structural alloys. In contrast to prior work that focuses on generating stable inorganic crystals, our approach targets the synthesizeability of a specific structural class: BCC/B2 superalloys, an underexplored family of materials with potential applications in extreme environments. Using three open-weight models (LLaMA-3.1, Gemma-2, and OLMo-2), we demonstrate that language models can be optimized for multiple design objectives using a single, unified reward signal through Direct Preference Optimization (DPO). Unlike prior approaches that rely on heuristic or human-in-the-loop feedback (costly), our reward signal is derived from thermodynamic phase calculations, offering a scientifically grounded criterion for model tuning. To our knowledge, this is the first demonstration of preference-tuning a language model using physics-grounded feedback for structural alloy design. The resulting framework is general and extensible, providing a path forward for intelligent design-space exploration across a range of physical science domains.

Peter Jansen, Peter Clark, Doug Downey et al.

Contemporary automated scientific discovery has focused on agents for generating scientific experiments, while systems that perform higher-level scientific activities such as theory building remain underexplored. In this work, we formulate the problem of synthesizing theories consisting of qualitative and quantitative laws from large corpora of scientific literature. We study theory generation at scale, using 13.7k source papers to synthesize 2.9k theories, examining how generation using literature-grounding versus parametric knowledge, and accuracy-focused versus novelty-focused generation objectives change theory properties. Our experiments show that, compared to using parametric LLM memory for generation, our literature-supported method creates theories that are significantly better at both matching existing evidence and at predicting future results from 4.6k subsequently-written papers

John Giorgi, Amanpreet Singh, Doug Downey et al. · allen-ai, uw

Topic pages aggregate useful information about an entity or concept into a single succinct and accessible article. Automated creation of topic pages would enable their rapid curation as information resources, providing an alternative to traditional web search. While most prior work has focused on generating topic pages about biographical entities, in this work, we develop a completely automated process to generate high-quality topic pages for scientific entities, with a focus on biomedical concepts. We release TOPICAL, a web app and associated open-source code, comprising a model pipeline combining retrieval, clustering, and prompting, that makes it easy for anyone to generate topic pages for a wide variety of biomedical entities on demand. In a human evaluation of 150 diverse topic pages generated using TOPICAL, we find that the vast majority were considered relevant, accurate, and coherent, with correct supporting citations. We make all code publicly available and host a free-to-use web app at: https://s2-topical.apps.allenai.org

Zejiang Shen, Kyle Lo, Lucy Lu Wang et al.

Accurately extracting structured content from PDFs is a critical first step for NLP over scientific papers. Recent work has improved extraction accuracy by incorporating elementary layout information, e.g., each token's 2D position on the page, into language model pretraining. We introduce new methods that explicitly model VIsual LAyout (VILA) groups, i.e., text lines or text blocks, to further improve performance. In our I-VILA approach, we show that simply inserting special tokens denoting layout group boundaries into model inputs can lead to a 1.9% Macro F1 improvement in token classification. In the H-VILA approach, we show that hierarchical encoding of layout-groups can result in up-to 47% inference time reduction with less than 0.8% Macro F1 loss. Unlike prior layout-aware approaches, our methods do not require expensive additional pretraining, only fine-tuning, which we show can reduce training cost by up to 95%. Experiments are conducted on a newly curated evaluation suite, S2-VLUE, that unifies existing automatically-labeled datasets and includes a new dataset of manual annotations covering diverse papers from 19 scientific disciplines. Pre-trained weights, benchmark datasets, and source code are available at https://github.com/allenai/VILA.

Daniel King, Doug Downey, Daniel S. Weld

Identification of new concepts in scientific literature can help power faceted search, scientific trend analysis, knowledge-base construction, and more, but current methods are lacking. Manual identification cannot keep up with the torrent of new publications, while the precision of existing automatic techniques is too low for many applications. We present an unsupervised concept extraction method for scientific literature that achieves much higher precision than previous work. Our approach relies on a simple but novel intuition: each scientific concept is likely to be introduced or popularized by a single paper that is disproportionately cited by subsequent papers mentioning the concept. From a corpus of computer science papers on arXiv, we find that our method achieves a Precision@1000 of 99%, compared to 86% for prior work, and a substantially better precision-yield trade-off across the top 15,000 extractions. To stimulate research in this area, we release our code and data (https://github.com/allenai/ForeCite).

Mike D'Arcy, Tom Hope, Larry Birnbaum et al.

We study the ability of LLMs to generate feedback for scientific papers and develop MARG, a feedback generation approach using multiple LLM instances that engage in internal discussion. By distributing paper text across agents, MARG can consume the full text of papers beyond the input length limitations of the base LLM, and by specializing agents and incorporating sub-tasks tailored to different comment types (experiments, clarity, impact) it improves the helpfulness and specificity of feedback. In a user study, baseline methods using GPT-4 were rated as producing generic or very generic comments more than half the time, and only 1.7 comments per paper were rated as good overall in the best baseline. Our system substantially improves the ability of GPT-4 to generate specific and helpful feedback, reducing the rate of generic comments from 60% to 29% and generating 3.7 good comments per paper (a 2.2x improvement).

Alan Li, Yixin Liu, Arpan Sarkar et al.

Scientific problem solving poses unique challenges for LLMs, requiring both deep domain knowledge and the ability to apply such knowledge through complex reasoning. While automated scientific reasoners hold great promise for assisting human scientists, there is currently no widely adopted holistic benchmark for evaluating scientific reasoning, and few approaches systematically disentangle the distinct roles of knowledge and reasoning in these tasks. To address these gaps, we introduce SciReas, a diverse suite of existing benchmarks for scientific reasoning tasks, and SciReas-Pro, a selective subset that requires more complex reasoning. Our holistic evaluation surfaces insights about scientific reasoning performance that remain hidden when relying on individual benchmarks alone. We then propose KRUX, a probing framework for studying the distinct roles of reasoning and knowledge in scientific tasks. Combining the two, we conduct an in-depth analysis that yields several key findings: (1) Retrieving task-relevant knowledge from model parameters is a critical bottleneck for LLMs in scientific reasoning; (2) Reasoning models consistently benefit from external knowledge added in-context on top of the reasoning enhancement; (3) Enhancing verbalized reasoning improves LLMs' ability to surface task-relevant knowledge. Finally, we conduct a lightweight analysis, comparing our science-focused data composition with concurrent efforts on long CoT SFT, and release SciLit01, a strong 8B baseline for scientific reasoning.

Vishakh Padmakumar, Joseph Chee Chang, Kyle Lo et al.

The increasing volume of academic literature makes it essential for researchers to organize, compare, and contrast collections of documents. Large language models (LLMs) can support this process by generating schemas defining shared aspects along which to compare papers. However, progress on schema generation has been slow due to: (i) ambiguity in reference-based evaluations, and (ii) lack of editing/refinement methods. Our work is the first to address both issues. First, we present an approach for augmenting unannotated table corpora with \emph{synthesized intents}, and apply it to create a dataset for studying schema generation conditioned on a given information need, thus reducing ambiguity. With this dataset, we show how incorporating table intents significantly improves baseline performance in reconstructing reference schemas. We start by comprehensively benchmarking several single-shot schema generation methods, including prompted LLM workflows and fine-tuned models, showing that smaller, open-weight models can be fine-tuned to be competitive with state-of-the-art prompted LLMs. Next, we propose several LLM-based schema refinement techniques and show that these can further improve schemas generated by these methods.

Jonathan Bragg, Mike D'Arcy, Nishant Balepur et al. · allen-ai

AI agents hold the potential to revolutionize scientific productivity by automating literature reviews, replicating experiments, analyzing data, and even proposing new directions of inquiry; indeed, there are now many such agents, ranging from general-purpose "deep research" systems to specialized science-specific agents, such as AI Scientist and AIGS. Rigorous evaluation of these agents is critical for progress. Yet existing benchmarks fall short on several fronts: they (1) fail to provide holistic, product-informed measures of real-world use cases such as science research; (2) lack reproducible agent tools necessary for a controlled comparison of core agentic capabilities; (3) do not account for confounding variables such as model cost and tool access; (4) do not provide standardized interfaces for quick agent prototyping and evaluation; and (5) lack comprehensive baseline agents necessary to identify true advances. In response, we define principles and tooling for more rigorously benchmarking agents. Using these, we present AstaBench, a suite that provides the first holistic measure of agentic ability to perform scientific research, comprising 2400+ problems spanning the entire scientific discovery process and multiple scientific domains, and including many problems inspired by actual user requests to deployed Asta agents. Our suite comes with the first scientific research environment with production-grade search tools that enable controlled, reproducible evaluation, better accounting for confounders. Alongside, we provide a comprehensive suite of nine science-optimized classes of Asta agents and numerous baselines. Our extensive evaluation of 57 agents across 22 agent classes reveals several interesting findings, most importantly that despite meaningful progress on certain individual aspects, AI remains far from solving the challenge of science research assistance.

David Wadden, Kejian Shi, Jacob Morrison et al.

We present SciRIFF (Scientific Resource for Instruction-Following and Finetuning), a dataset of 137K instruction-following instances for training and evaluation, covering 54 tasks. These tasks span five core scientific literature understanding capabilities: information extraction, summarization, question answering, claim verification, and classification. SciRIFF is unique in being entirely expert-written, high-quality instruction-following dataset for extracting and synthesizing information from research literature across diverse scientific fields. It features complex instructions with long input contexts, detailed task descriptions, and structured outputs. To demonstrate its utility, we finetune a series of large language models (LLMs) using a mix of general-domain and SciRIFF instructions. On nine out-of-distribution held-out tasks (referred to as SciRIFF-Eval), LLMs finetuned on SciRIFF achieve 70.6% average improvement over baselines trained only on general-domain instructions. SciRIFF facilitates the development and evaluation of LLMs to help researchers navigate the rapidly growing body of scientific literature.

Qingyun Wang, Doug Downey, Heng Ji et al.

We explore and enhance the ability of neural language models to generate novel scientific directions grounded in literature. Work on literature-based hypothesis generation has traditionally focused on binary link prediction--severely limiting the expressivity of hypotheses. This line of work also does not focus on optimizing novelty. We take a dramatic departure with a novel setting in which models use as input background contexts (e.g., problems, experimental settings, goals), and output natural language ideas grounded in literature. We present SciMON, a modeling framework that uses retrieval of "inspirations" from past scientific papers, and explicitly optimizes for novelty by iteratively comparing to prior papers and updating idea suggestions until sufficient novelty is achieved. Comprehensive evaluations reveal that GPT-4 tends to generate ideas with overall low technical depth and novelty, while our methods partially mitigate this issue. Our work represents a first step toward evaluating and developing language models that generate new ideas derived from the scientific literature

Ana Marasović, Iz Beltagy, Doug Downey et al.

Self-rationalization models that predict task labels and generate free-text elaborations for their predictions could enable more intuitive interaction with NLP systems. These models are, however, currently trained with a large amount of human-written free-text explanations for each task which hinders their broader usage. We propose to study a more realistic setting of self-rationalization using few training examples. We present FEB -- a standardized collection of four existing English-language datasets and associated metrics. We identify the right prompting approach by extensively exploring natural language prompts on FEB. Then, by using this prompt and scaling the model size, we demonstrate that making progress on few-shot self-rationalization is possible. We show there is still ample room for improvement in this task: the average plausibility of generated explanations assessed by human annotators is at most 51% (with GPT-3), while plausibility of human explanations is 76%. We hope that FEB and our proposed approach will spur the community to take on the few-shot self-rationalization challenge.

Marissa Radensky, Doug Downey, Kyle Lo et al.

Explanations are well-known to improve recommender systems' transparency. These explanations may be local, explaining an individual recommendation, or global, explaining the recommender model in general. Despite their widespread use, there has been little investigation into the relative benefits of these two approaches. Do they provide the same benefits to users, or do they serve different purposes? We conducted a 30-participant exploratory study and a 30-participant controlled user study with a research-paper recommender system to analyze how providing participants local, global, or both explanations influences user understanding of system behavior. Our results provide evidence suggesting that both explanations are more helpful than either alone for explaining how to improve recommendations, yet both appeared less helpful than global alone for efficiency in identifying false positives and negatives. However, we note that the two explanation approaches may be better compared in the context of a higher-stakes or more opaque domain.

Victor S. Bursztyn, Jennifer Healey, Nedim Lipka et al.

Conversations aimed at determining good recommendations are iterative in nature. People often express their preferences in terms of a critique of the current recommendation (e.g., "It doesn't look good for a date"), requiring some degree of common sense for a preference to be inferred. In this work, we present a method for transforming a user critique into a positive preference (e.g., "I prefer more romantic") in order to retrieve reviews pertaining to potentially better recommendations (e.g., "Perfect for a romantic dinner"). We leverage a large neural language model (LM) in a few-shot setting to perform critique-to-preference transformation, and we test two methods for retrieving recommendations: one that matches embeddings, and another that fine-tunes an LM for the task. We instantiate this approach in the restaurant domain and evaluate it using a new dataset of restaurant critiques. In an ablation study, we show that utilizing critique-to-preference transformation improves recommendations, and that there are at least three general cases that explain this improved performance.

Arie Cattan, Sophie Johnson, Daniel Weld et al.

Determining coreference of concept mentions across multiple documents is a fundamental task in natural language understanding. Previous work on cross-document coreference resolution (CDCR) typically considers mentions of events in the news, which seldom involve abstract technical concepts that are prevalent in science and technology. These complex concepts take diverse or ambiguous forms and have many hierarchical levels of granularity (e.g., tasks and subtasks), posing challenges for CDCR. We present a new task of Hierarchical CDCR (H-CDCR) with the goal of jointly inferring coreference clusters and hierarchy between them. We create SciCo, an expert-annotated dataset for H-CDCR in scientific papers, 3X larger than the prominent ECB+ resource. We study strong baseline models that we customize for H-CDCR, and highlight challenges for future work.

Sean MacAvaney, Andrew Yates, Sergey Feldman et al.

Managing the data for Information Retrieval (IR) experiments can be challenging. Dataset documentation is scattered across the Internet and once one obtains a copy of the data, there are numerous different data formats to work with. Even basic formats can have subtle dataset-specific nuances that need to be considered for proper use. To help mitigate these challenges, we introduce a new robust and lightweight tool (ir_datasets) for acquiring, managing, and performing typical operations over datasets used in IR. We primarily focus on textual datasets used for ad-hoc search. This tool provides both a Python and command line interface to numerous IR datasets and benchmarks. To our knowledge, this is the most extensive tool of its kind. Integrations with popular IR indexing and experimentation toolkits demonstrate the tool's utility. We also provide documentation of these datasets through the ir_datasets catalog: https://ir-datasets.com/. The catalog acts as a hub for information on datasets used in IR, providing core information about what data each benchmark provides as well as links to more detailed information. We welcome community contributions and intend to continue to maintain and grow this tool.

Sean MacAvaney, Sergey Feldman, Nazli Goharian et al.

Pretrained contextualized language models such as BERT and T5 have established a new state-of-the-art for ad-hoc search. However, it is not yet well-understood why these methods are so effective, what makes some variants more effective than others, and what pitfalls they may have. We present a new comprehensive framework for Analyzing the Behavior of Neural IR ModeLs (ABNIRML), which includes new types of diagnostic probes that allow us to test several characteristics -- such as writing styles, factuality, sensitivity to paraphrasing and word order -- that are not addressed by previous techniques. To demonstrate the value of the framework, we conduct an extensive empirical study that yields insights into the factors that contribute to the neural model's gains, and identify potential unintended biases the models exhibit. Some of our results confirm conventional wisdom, like that recent neural ranking models rely less on exact term overlap with the query, and instead leverage richer linguistic information, evidenced by their higher sensitivity to word and sentence order. Other results are more surprising, such as that some models (e.g., T5 and ColBERT) are biased towards factually correct (rather than simply relevant) texts. Further, some characteristics vary even for the same base language model, and other characteristics can appear due to random variations during model training.

David Demeter, Gregory Kimmel, Doug Downey

Neural Network Language Models (NNLMs) generate probability distributions by applying a softmax function to a distance metric formed by taking the dot product of a prediction vector with all word vectors in a high-dimensional embedding space. The dot-product distance metric forms part of the inductive bias of NNLMs. Although NNLMs optimize well with this inductive bias, we show that this results in a sub-optimal ordering of the embedding space that structurally impoverishes some words at the expense of others when assigning probability. We present numerical, theoretical and empirical analyses showing that words on the interior of the convex hull in the embedding space have their probability bounded by the probabilities of the words on the hull.

Yiben Yang, Chaitanya Malaviya, Jared Fernandez et al.

Recent advances in commonsense reasoning depend on large-scale human-annotated training data to achieve peak performance. However, manual curation of training examples is expensive and has been shown to introduce annotation artifacts that neural models can readily exploit and overfit on. We investigate G-DAUG^C, a novel generative data augmentation method that aims to achieve more accurate and robust learning in the low-resource setting. Our approach generates synthetic examples using pretrained language models, and selects the most informative and diverse set of examples for data augmentation. In experiments with multiple commonsense reasoning benchmarks, G-DAUG^C consistently outperforms existing data augmentation methods based on back-translation, and establishes a new state-of-the-art on WinoGrande, CODAH, and CommonsenseQA. Further, in addition to improvements in in-distribution accuracy, G-DAUG^C-augmented training also enhances out-of-distribution generalization, showing greater robustness against adversarial or perturbed examples. Our analysis demonstrates that G-DAUG^C produces a diverse set of fluent training examples, and that its selection and training approaches are important for performance. Our findings encourage future research toward generative data augmentation to enhance both in-distribution learning and out-of-distribution generalization.

Suchin Gururangan, Ana Marasović, Swabha Swayamdipta et al.

Language models pretrained on text from a wide variety of sources form the foundation of today's NLP. In light of the success of these broad-coverage models, we investigate whether it is still helpful to tailor a pretrained model to the domain of a target task. We present a study across four domains (biomedical and computer science publications, news, and reviews) and eight classification tasks, showing that a second phase of pretraining in-domain (domain-adaptive pretraining) leads to performance gains, under both high- and low-resource settings. Moreover, adapting to the task's unlabeled data (task-adaptive pretraining) improves performance even after domain-adaptive pretraining. Finally, we show that adapting to a task corpus augmented using simple data selection strategies is an effective alternative, especially when resources for domain-adaptive pretraining might be unavailable. Overall, we consistently find that multi-phase adaptive pretraining offers large gains in task performance.

Arman Cohan, Sergey Feldman, Iz Beltagy et al.

Representation learning is a critical ingredient for natural language processing systems. Recent Transformer language models like BERT learn powerful textual representations, but these models are targeted towards token- and sentence-level training objectives and do not leverage information on inter-document relatedness, which limits their document-level representation power. For applications on scientific documents, such as classification and recommendation, the embeddings power strong performance on end tasks. We propose SPECTER, a new method to generate document-level embedding of scientific documents based on pretraining a Transformer language model on a powerful signal of document-level relatedness: the citation graph. Unlike existing pretrained language models, SPECTER can be easily applied to downstream applications without task-specific fine-tuning. Additionally, to encourage further research on document-level models, we introduce SciDocs, a new evaluation benchmark consisting of seven document-level tasks ranging from citation prediction, to document classification and recommendation. We show that SPECTER outperforms a variety of competitive baselines on the benchmark.

Benjamin Charles Germain Lee, Doug Downey, Kyle Lo et al.

Research in human-centered AI has shown the benefits of systems that can explain their predictions. Methods that allow an AI to take advice from humans in response to explanations are similarly useful. While both capabilities are well-developed for transparent learning models (e.g., linear models and GA$^2$Ms), and recent techniques (e.g., LIME and SHAP) can generate explanations for opaque models, little attention has been given to advice methods for opaque models. This paper introduces LIMEADE, the first general framework that translates both positive and negative advice (expressed using high-level vocabulary such as that employed by post-hoc explanations) into an update to an arbitrary, underlying opaque model. We demonstrate the generality of our approach with case studies on seventy real-world models across two broad domains: image classification and text recommendation. We show our method improves accuracy compared to a rigorous baseline on the image classification domains. For the text modality, we apply our framework to a neural recommender system for scientific papers on a public website; our user study shows that our framework leads to significantly higher perceived user control, trust, and satisfaction.

David Demeter, Doug Downey

Neural network language models (NNLMs) have achieved ever-improving accuracy due to more sophisticated architectures and increasing amounts of training data. However, the inductive bias of these models (formed by the distributional hypothesis of language), while ideally suited to modeling most running text, results in key limitations for today's models. In particular, the models often struggle to learn certain spatial, temporal, or quantitative relationships, which are commonplace in text and are second-nature for human readers. Yet, in many cases, these relationships can be encoded with simple mathematical or logical expressions. How can we augment today's neural models with such encodings? In this paper, we propose a general methodology to enhance the inductive bias of NNLMs by incorporating simple functions into a neural architecture to form a hierarchical neural-symbolic language model (NSLM). These functions explicitly encode symbolic deterministic relationships to form probability distributions over words. We explore the effectiveness of this approach on numbers and geographic locations, and show that NSLMs significantly reduce perplexity in small-corpus language modeling, and that the performance improvement persists for rare tokens even on much larger corpora. The approach is simple and general, and we discuss how it can be applied to other word classes beyond numbers and geography.

Ruimin Zhu, Thanapon Noraset, Alisa Liu et al.

Word embeddings capture syntactic and semantic information about words. Definition modeling aims to make the semantic content in each embedding explicit, by outputting a natural language definition based on the embedding. However, existing definition models are limited in their ability to generate accurate definitions for different senses of the same word. In this paper, we introduce a new method that enables definition modeling for multiple senses. We show how a Gumble-Softmax approach outperforms baselines at matching sense-specific embeddings to definitions during training. In experiments, our multi-sense definition model improves recall over a state-of-the-art single-sense definition model by a factor of three, without harming precision.

Chandra Bhagavatula, Ronan Le Bras, Chaitanya Malaviya et al.

Abductive reasoning is inference to the most plausible explanation. For example, if Jenny finds her house in a mess when she returns from work, and remembers that she left a window open, she can hypothesize that a thief broke into her house and caused the mess, as the most plausible explanation. While abduction has long been considered to be at the core of how people interpret and read between the lines in natural language (Hobbs et al., 1988), there has been relatively little research in support of abductive natural language inference and generation. We present the first study that investigates the viability of language-based abductive reasoning. We introduce a challenge dataset, ART, that consists of over 20k commonsense narrative contexts and 200k explanations. Based on this dataset, we conceptualize two new tasks -- (i) Abductive NLI: a multiple-choice question answering task for choosing the more likely explanation, and (ii) Abductive NLG: a conditional generation task for explaining given observations in natural language. On Abductive NLI, the best model achieves 68.9% accuracy, well below human performance of 91.4%. On Abductive NLG, the current best language generators struggle even more, as they lack reasoning capabilities that are trivial for humans. Our analysis leads to new insights into the types of reasoning that deep pre-trained language models fail to perform--despite their strong performance on the related but more narrowly defined task of entailment NLI--pointing to interesting avenues for future research.