Mausam

Pranjal Aggarwal, Aman Madaan, Ankit Anand et al. · cmu

Large language models (LLMs) are now available from cloud API providers in various sizes and configurations. While this diversity offers a broad spectrum of choices, effectively leveraging the options to optimize computational cost and performance remains challenging. In this work, we present Automix, an approach that strategically routes queries to larger LMs, based on the approximate correctness of outputs from a smaller LM. Central to Automix are two key technical contributions. First, it has a few-shot self-verification mechanism, which estimates the reliability of its own outputs without requiring extensive training. Second, given that self-verification can be noisy, it employs a POMDP based router that can effectively select an appropriately sized model, based on answer confidence. Experiments across five language models and five challenging datasets show that Automix consistently surpasses strong baselines, reducing computational cost by over 50% for comparable performance.

Keshav Kolluru, Gabriel Stanovsky, Mausam

Proper noun compounds, e.g., "Covid vaccine", convey information in a succinct manner (a "Covid vaccine" is a "vaccine that immunizes against the Covid disease"). These are commonly used in short-form domains, such as news headlines, but are largely ignored in information-seeking applications. To address this limitation, we release a new manually annotated dataset, ProNCI, consisting of 22.5K proper noun compounds along with their free-form semantic interpretations. ProNCI is 60 times larger than prior noun compound datasets and also includes non-compositional examples, which have not been previously explored. We experiment with various neural models for automatically generating the semantic interpretations from proper noun compounds, ranging from few-shot prompting to supervised learning, with varying degrees of knowledge about the constituent nouns. We find that adding targeted knowledge, particularly about the common noun, results in performance gains of upto 2.8%. Finally, we integrate our model generated interpretations with an existing Open IE system and observe an 7.5% increase in yield at a precision of 85%. The dataset and code are available at https://github.com/dair-iitd/pronci.

Aayush Kumar Tyagi, Chirag Mohapatra, Prasenjit Das et al.

Multi-class cell detection and counting is an essential task for many pathological diagnoses. Manual counting is tedious and often leads to inter-observer variations among pathologists. While there exist multiple, general-purpose, deep learning-based object detection and counting methods, they may not readily transfer to detecting and counting cells in medical images, due to the limited data, presence of tiny overlapping objects, multiple cell types, severe class-imbalance, minute differences in size/shape of cells, etc. In response, we propose guided posterior regularization (DeGPR), which assists an object detector by guiding it to exploit discriminative features among cells. The features may be pathologist-provided or inferred directly from visual data. We validate our model on two publicly available datasets (CoNSeP and MoNuSAC), and on MuCeD, a novel dataset that we contribute. MuCeD consists of 55 biopsy images of the human duodenum for predicting celiac disease. We perform extensive experimentation with three object detection baselines on three datasets to show that DeGPR is model-agnostic, and consistently improves baselines obtaining up to 9% (absolute) mAP gains.

Yatin Nandwani, Rishabh Ranjan, Mausam et al.

There is a recent focus on designing architectures that have an Integer Linear Programming (ILP) layer within a neural model (referred to as Neural ILP in this paper). Neural ILP architectures are suitable for pure reasoning tasks that require data-driven constraint learning or for tasks requiring both perception (neural) and reasoning (ILP). A recent SOTA approach for end-to-end training of Neural ILP explicitly defines gradients through the ILP black box (Paulus et al. 2021) - this trains extremely slowly, owing to a call to the underlying ILP solver for every training data point in a minibatch. In response, we present an alternative training strategy that is solver-free, i.e., does not call the ILP solver at all at training time. Neural ILP has a set of trainable hyperplanes (for cost and constraints in ILP), together representing a polyhedron. Our key idea is that the training loss should impose that the final polyhedron separates the positives (all constraints satisfied) from the negatives (at least one violated constraint or a suboptimal cost value), via a soft-margin formulation. While positive example(s) are provided as part of the training data, we devise novel techniques for generating negative samples. Our solution is flexible enough to handle equality as well as inequality constraints. Experiments on several problems, both perceptual as well as symbolic, which require learning the constraints of an ILP, show that our approach has superior performance and scales much better compared to purely neural baselines and other state-of-the-art models that require solver-based training. In particular, we are able to obtain excellent performance in 9 x 9 symbolic and visual sudoku, to which the other Neural ILP solver is not able to scale.

Mohd Zaki, Jayadeva, Mausam et al.

Information extraction and textual comprehension from materials literature are vital for developing an exhaustive knowledge base that enables accelerated materials discovery. Language models have demonstrated their capability to answer domain-specific questions and retrieve information from knowledge bases. However, there are no benchmark datasets in the materials domain that can evaluate the understanding of the key concepts by these language models. In this work, we curate a dataset of 650 challenging questions from the materials domain that require the knowledge and skills of a materials student who has cleared their undergraduate degree. We classify these questions based on their structure and the materials science domain-based subcategories. Further, we evaluate the performance of GPT-3.5 and GPT-4 models on solving these questions via zero-shot and chain of thought prompting. It is observed that GPT-4 gives the best performance (~62% accuracy) as compared to GPT-3.5. Interestingly, in contrast to the general observation, no significant improvement in accuracy is observed with the chain of thought prompting. To evaluate the limitations, we performed an error analysis, which revealed conceptual errors (~64%) as the major contributor compared to computational errors (~36%) towards the reduced performance of LLMs. We hope that the dataset and analysis performed in this work will promote further research in developing better materials science domain-specific LLMs and strategies for information extraction.

Mayur Patidar, Prayushi Faldu, Avinash Singh et al.

When answering natural language questions over knowledge bases, missing facts, incomplete schema and limited scope naturally lead to many questions being unanswerable. While answerability has been explored in other QA settings, it has not been studied for QA over knowledge bases (KBQA). We create GrailQAbility, a new benchmark KBQA dataset with unanswerability, by first identifying various forms of KB incompleteness that make questions unanswerable, and then systematically adapting GrailQA (a popular KBQA dataset with only answerable questions). Experimenting with three state-of-the-art KBQA models, we find that all three models suffer a drop in performance even after suitable adaptation for unanswerable questions. In addition, these often detect unanswerability for wrong reasons and find specific forms of unanswerability particularly difficult to handle. This underscores the need for further research in making KBQA systems robust to unanswerability

Shubham Mittal, Keshav Kolluru, Soumen Chakrabarti et al.

Automated completion of open knowledge bases (Open KBs), which are constructed from triples of the form (subject phrase, relation phrase, object phrase), obtained via open information extraction (Open IE) system, are useful for discovering novel facts that may not be directly present in the text. However, research in Open KB completion (Open KBC) has so far been limited to resource-rich languages like English. Using the latest advances in multilingual Open IE, we construct the first multilingual Open KBC dataset, called mOKB6, containing facts from Wikipedia in six languages (including English). Improving the previous Open KB construction pipeline by doing multilingual coreference resolution and keeping only entity-linked triples, we create a dense Open KB. We experiment with several models for the task and observe a consistent benefit of combining languages with the help of shared embedding space as well as translations of facts. We also observe that current multilingual models struggle to remember facts seen in languages of different scripts.

Shreya Sharma, Jigyasa Gupta, Shreshth Tuli et al.

Our goal is to enable a robot to learn how to sequence its actions to perform tasks specified as natural language instructions, given successful demonstrations from a human partner. The ability to plan high-level tasks can be factored as (i) inferring specific goal predicates that characterize the task implied by a language instruction for a given world state and (ii) synthesizing a feasible goal-reaching action-sequence with such predicates. For the former, we leverage a neural network prediction model, while utilizing a symbolic planner for the latter. We introduce a novel neuro-symbolic model, GoalNet, for contextual and task dependent inference of goal predicates from human demonstrations and linguistic task descriptions. GoalNet combines (i) learning, where dense representations are acquired for language instruction and the world state that enables generalization to novel settings and (ii) planning, where the cause-effect modeling by the symbolic planner eschews irrelevant predicates facilitating multi-stage decision making in large domains. GoalNet demonstrates a significant improvement (51%) in the task completion rate in comparison to a state-of-the-art rule-based approach on a benchmark data set displaying linguistic variations, particularly for multi-stage instructions.

Vipul Rathore, Rajdeep Dhingra, Parag Singla et al.

We tackle the problem of zero-shot cross-lingual transfer in NLP tasks via the use of language adapters (LAs). Most of the earlier works have explored training with adapter of a single source (often English), and testing either using the target LA or LA of another related language. Training target LA requires unlabeled data, which may not be readily available for low resource unseen languages: those that are neither seen by the underlying multilingual language model (e.g., mBERT), nor do we have any (labeled or unlabeled) data for them. We posit that for more effective cross-lingual transfer, instead of just one source LA, we need to leverage LAs of multiple (linguistically or geographically related) source languages, both at train and test-time - which we investigate via our novel neural architecture, ZGUL. Extensive experimentation across four language groups, covering 15 unseen target languages, demonstrates improvements of up to 3.2 average F1 points over standard fine-tuning and other strong baselines on POS tagging and NER tasks. We also extend ZGUL to settings where either (1) some unlabeled data or (2) few-shot training examples are available for the target language. We find that ZGUL continues to outperform baselines in these settings too.

Tanishq Gupta, Mohd Zaki, Devanshi Khatsuriya et al.

A crucial component in the curation of KB for a scientific domain (e.g., materials science, foods & nutrition, fuels) is information extraction from tables in the domain's published research articles. To facilitate research in this direction, we define a novel NLP task of extracting compositions of materials (e.g., glasses) from tables in materials science papers. The task involves solving several challenges in concert, such as tables that mention compositions have highly varying structures; text in captions and full paper needs to be incorporated along with data in tables; and regular languages for numbers, chemical compounds and composition expressions must be integrated into the model. We release a training dataset comprising 4,408 distantly supervised tables, along with 1,475 manually annotated dev and test tables. We also present a strong baseline DISCOMAT, that combines multiple graph neural networks with several task-specific regular expressions, features, and constraints. We show that DISCOMAT outperforms recent table processing architectures by significant margins.

Kausik Hira, Mohd Zaki, Dhruvil Sheth et al.

The discovery of new materials has a documented history of propelling human progress for centuries and more. The behaviour of a material is a function of its composition, structure, and properties, which further depend on its processing and testing conditions. Recent developments in deep learning and natural language processing have enabled information extraction at scale from published literature such as peer-reviewed publications, books, and patents. However, this information is spread in multiple formats, such as tables, text, and images, and with little or no uniformity in reporting style giving rise to several machine learning challenges. Here, we discuss, quantify, and document these challenges in automated information extraction (IE) from materials science literature towards the creation of a large materials science knowledge base. Specifically, we focus on IE from text and tables and outline several challenges with examples. We hope the present work inspires researchers to address the challenges in a coherent fashion, providing a fillip to IE towards developing a materials knowledge base.

Ananjan Nandi, Navdeep Kaur, Parag Singla et al.

High-quality and high-coverage rule sets are imperative to the success of Neuro-Symbolic Knowledge Graph Completion (NS-KGC) models, because they form the basis of all symbolic inferences. Recent literature builds neural models for generating rule sets, however, preliminary experiments show that they struggle with maintaining high coverage. In this work, we suggest three simple augmentations to existing rule sets: (1) transforming rules to their abductive forms, (2) generating equivalent rules that use inverse forms of constituent relations and (3) random walks that propose new rules. Finally, we prune potentially low quality rules. Experiments over four datasets and five ruleset-baseline settings suggest that these simple augmentations consistently improve results, and obtain up to 7.1 pt MRR and 8.5 pt Hits@1 gains over using rules without augmentations.

Ananjan Nandi, Navdeep Kaur, Parag Singla et al.

We consider two popular approaches to Knowledge Graph Completion (KGC): textual models that rely on textual entity descriptions, and structure-based models that exploit the connectivity structure of the Knowledge Graph (KG). Preliminary experiments show that these approaches have complementary strengths: structure-based models perform exceptionally well when the gold answer is easily reachable from the query head in the KG, while textual models exploit descriptions to give good performance even when the gold answer is not easily reachable. In response, we propose DynaSemble, a novel method for learning query-dependent ensemble weights to combine these approaches by using the distributions of scores assigned by the models in the ensemble to all candidate entities. DynaSemble achieves state-of-the-art results on three standard KGC datasets, with up to 6.8 pt MRR and 8.3 pt Hits@1 gains over the best baseline model for the WN18RR dataset.

Anand Brahmbhatt, Vipul Rathore, Mausam et al.

Recent literature has seen a significant focus on building machine learning models with specific properties such as fairness, i.e., being non-biased with respect to a given set of attributes, calibration i.e., model confidence being aligned with its predictive accuracy, and explainability, i.e., ability to be understandable to humans. While there has been work focusing on each of these aspects individually, researchers have shied away from simultaneously addressing more than one of these dimensions. In this work, we address the problem of building models which are both fair and calibrated. We work with a specific definition of fairness, which closely matches [Biswas et. al. 2019], and has the nice property that Bayes optimal classifier has the maximum possible fairness under our definition. We show that an existing negative result towards achieving a fair and calibrated model [Kleinberg et. al. 2017] does not hold for our definition of fairness. Further, we show that ensuring group-wise calibration with respect to the sensitive attributes automatically results in a fair model under our definition. Using this result, we provide a first cut approach for achieving fair and calibrated models, via a simple post-processing technique based on temperature scaling. We then propose modifications of existing calibration losses to perform group-wise calibration, as a way of achieving fair and calibrated models in a variety of settings. Finally, we perform extensive experimentation of these techniques on a diverse benchmark of datasets, and present insights on the pareto-optimality of the resulting solutions.

Mayur Patidar, Riya Sawhney, Avinash Singh et al.

Existing Knowledge Base Question Answering (KBQA) architectures are hungry for annotated data, which make them costly and time-consuming to deploy. We introduce the problem of few-shot transfer learning for KBQA, where the target domain offers only a few labeled examples, but a large labeled training dataset is available in a source domain. We propose a novel KBQA architecture called FuSIC-KBQA that performs KB-retrieval using multiple source-trained retrievers, re-ranks using an LLM and uses this as input for LLM few-shot in-context learning to generate logical forms. These are further refined using execution-guided feedback. Experiments over multiple source-target KBQA pairs of varying complexity show that FuSIC-KBQA significantly outperforms adaptations of SoTA KBQA models for this setting. Additional experiments show that FuSIC-KBQA also outperforms SoTA KBQA models in the in-domain setting when training data is limited.

Harshvardhan Srivastava, Kanav Pruthi, Soumen Chakrabarti et al.

End-to-end conversational recommendation systems (CRS) generate responses by leveraging both dialog history and a knowledge base (KB). A CRS mainly faces three key challenges: (1) at each turn, it must decide if recommending a KB entity is appropriate; if so, it must identify the most relevant KB entity to recommend; and finally, it must recommend the entity in a fluent utterance that is consistent with the conversation history. Recent CRSs do not pay sufficient attention to these desiderata, often generating unfluent responses or not recommending (relevant) entities at the right turn. We introduce a new CRS we call CoRE-CoG. CoRE-CoG addresses the limitations in prior systems by implementing (1) a recommendation trigger that decides if the system utterance should include an entity, (2) a type pruning module that improves the relevance of recommended entities, and (3) a novel constrained response generator to make recommendations while maintaining fluency. Together, these modules ensure simultaneous accurate recommendation decisions and fluent system utterances. Experiments with recent benchmarks show the superiority particularly on conditional generation sub-tasks with close to 10 F1 and 4 Recall@1 percent points gain over baselines.

Aayush Kumar Tyagi, Vaibhav Mishra, Prathosh A. P. et al.

Cell segmentation in histopathological images is vital for diagnosis, and treatment of several diseases. Annotating data is tedious, and requires medical expertise, making it difficult to employ supervised learning. Instead, we study a weakly supervised setting, where only bounding box supervision is available, and present the use of Segment Anything (SAM) for this without any finetuning, i.e., directly utilizing the pre-trained model. We propose BoxCell, a cell segmentation framework that utilizes SAM's capability to interpret bounding boxes as prompts, \emph{both} at train and test times. At train time, gold bounding boxes given to SAM produce (pseudo-)masks, which are used to train a standalone segmenter. At test time, BoxCell generates two segmentation masks: (1) generated by this standalone segmenter, and (2) a trained object detector outputs bounding boxes, which are given as prompts to SAM to produce another mask. Recognizing complementary strengths, we reconcile the two segmentation masks using a novel integer programming formulation with intensity and spatial constraints. We experiment on three publicly available cell segmentation datasets namely, CoNSep, MoNuSeg, and TNBC, and find that BoxCell significantly outperforms existing box supervised image segmentation models, obtaining 6-10 point Dice gains.

Samrajnee Ghosh, Naman Agarwal, Hemanshu Garg et al.

Cognitive science research treats visual perception, the ability to understand and make sense of a visual input, as one of the early developmental signs of intelligence. Its TVPS-4 framework categorizes and tests human perception into seven skills such as visual discrimination, and form constancy. Do Multimodal Large Language Models (MLLMs) match up to humans in basic perception? Even though there are many benchmarks that evaluate MLLMs on advanced reasoning and knowledge skills, there is limited research that focuses evaluation on simple perception. In response, we introduce Percept-V, a dataset containing 6000 program-generated uncontaminated images divided into 30 domains, where each domain tests one or more TVPS-4 skills. Our focus is on perception, so we make our domains quite simple and the reasoning and knowledge required for solving them are minimal. Since modern-day MLLMs can solve much more complex tasks, our a-priori expectation is that they will solve these domains very easily. Contrary to our belief, our experiments show a weak performance of SoTA proprietary and open-source MLLMs compared to very high human performance on Percept-V. We find that as number of objects in the image increases, performance goes down rather fast. Our experiments also identify the perception skills that are considerably harder for all models.

Aayush Kumar Tyagi, Vaibhav Mishra, Ashok Tiwari et al.

Celiac disease is an autoimmune disorder triggered by the consumption of gluten. It causes damage to the villi, the finger-like projections in the small intestine that are responsible for nutrient absorption. Additionally, the crypts, which form the base of the villi, are also affected, impairing the regenerative process. The deterioration in villi length, computed as the villi-to-crypt length ratio, indicates the severity of celiac disease. However, manual measurement of villi-crypt length can be both time-consuming and susceptible to inter-observer variability, leading to inconsistencies in diagnosis. While some methods can perform measurement as a post-hoc process, they are prone to errors in the initial stages. This gap underscores the need for pathologically driven solutions that enhance measurement accuracy and reduce human error in celiac disease assessments. Our proposed method, MeasureNet, is a pathologically driven polyline detection framework incorporating polyline localization and object-driven losses specifically designed for measurement tasks. Furthermore, we leverage segmentation model to provide auxiliary guidance about crypt location when crypt are partially visible. To ensure that model is not overdependent on segmentation mask we enhance model robustness through a mask feature mixup technique. Additionally, we introduce a novel dataset for grading celiac disease, consisting of 750 annotated duodenum biopsy images. MeasureNet achieves an 82.66% classification accuracy for binary classification and 81% accuracy for multi-class grading of celiac disease. Code: https://github.com/dair-iitd/MeasureNet

Daman Arora, Himanshu Gaurav Singh, Mausam

The performance of large language models (LLMs) on existing reasoning benchmarks has significantly improved over the past years. In response, we present JEEBench, a considerably more challenging benchmark dataset for evaluating the problem solving abilities of LLMs. We curate 515 challenging pre-engineering mathematics, physics and chemistry problems from the highly competitive IIT JEE-Advanced exam. Long-horizon reasoning on top of deep in-domain knowledge is essential for solving problems in this benchmark. Our evaluation on various open-source and proprietary models reveals that the highest performance, even after using techniques like self-consistency, self-refinement and chain-of-thought prompting, is less than 40%. The typical failure modes of GPT-4, the best model, are errors in algebraic manipulation, difficulty in grounding abstract concepts into mathematical equations accurately and failure in retrieving relevant domain-specific concepts. We also observe that by mere prompting, GPT-4 is unable to assess risk introduced by negative marking for incorrect answers. For this, we develop a post-hoc confidence-thresholding method over self-consistency, which enables effective response selection. We hope that our challenging benchmark will guide future re-search in problem-solving using LLMs.

Harshit Bisht, Vinay Kumar, Kevin Maik Jablonka et al.

A growing body of work pursues AI scientists capable of end-to-end autonomous scientific discovery. This position paper argues that although they already function as co-scientists, agentic AI scientists are not built for autonomous scientific discovery. We identify the following challenges in building and deploying autonomous AI scientists: (1) Problem selection is influenced by the McNamara fallacy; (2) Agents are built on large language models (LLMs) whose training corpora omit tacit procedural and failure knowledge of laboratory practice; (3) Preference optimisation during post-training compresses output diversity toward consensus; and (4) Most scientific benchmarks measure single-turn prediction accuracy and lack feedback from physical experiments back to the computational model. These challenges are not just questions of scale and scaffolding; they require revisiting fundamental design choices. To build truly autonomous AI scientists, we recommend the use of scientific simulations as verifiers for training, the design of persistent world models that represent the shifting objectives governing real investigations, the establishment of a centralized preregistration repository for all AI-generated hypotheses, and application driven by scientific need rather than tool affordance.

Vinay Kumar, Satyendra Rajput, Mausam et al.

The promise of AI-driven scientific discovery hinges on whether AI agents can autonomously design and execute the computational workflows that underpin modern science. Molecular dynamics (MD) simulation presents a natural test bed to stress-test this claim; it requires translating physical intuition into syntactically and semantically correct input scripts, reasoning about initial and boundary conditions, diagnosing numerically unstable trajectories, and interpreting outputs against known physical behavior and laws. We introduce MDGYM, a benchmark of 169 expert-curated MD simulations spanning LAMMPS and GROMACS, two widely used MD packages, across three increasing difficulty levels. We evaluate three agentic frameworks -- Claude Code, Codex, and OpenHands -- with four LLMs, and find that all perform poorly: even the strongest agent solves only 21\% of easy-level tasks, with less than 10\% at higher difficulties. Trajectory analysis reveals a characteristic pattern of failure -- agents successfully invoke simulation machinery but produce physically unstable configurations, fabricate numerical outputs without executing the underlying computation, or abandon tasks prematurely rather than iterating through simulation-specific errors. These failure modes are qualitatively distinct from those observed in general software engineering benchmarks, indicating that fluent code generation does not transfer to grounded physical reasoning.

Ilia Kuznetsov, Osama Mohammed Afzal, Koen Dercksen et al.

The number of scientific articles produced every year is growing rapidly. Providing quality control over them is crucial for scientists and, ultimately, for the public good. In modern science, this process is largely delegated to peer review -- a distributed procedure in which each submission is evaluated by several independent experts in the field. Peer review is widely used, yet it is hard, time-consuming, and prone to error. Since the artifacts involved in peer review -- manuscripts, reviews, discussions -- are largely text-based, Natural Language Processing has great potential to improve reviewing. As the emergence of large language models (LLMs) has enabled NLP assistance for many new tasks, the discussion on machine-assisted peer review is picking up the pace. Yet, where exactly is help needed, where can NLP help, and where should it stand aside? The goal of our paper is to provide a foundation for the future efforts in NLP for peer-reviewing assistance. We discuss peer review as a general process, exemplified by reviewing at AI conferences. We detail each step of the process from manuscript submission to camera-ready revision, and discuss the associated challenges and opportunities for NLP assistance, illustrated by existing work. We then turn to the big challenges in NLP for peer review as a whole, including data acquisition and licensing, operationalization and experimentation, and ethical issues. To help consolidate community efforts, we create a companion repository that aggregates key datasets pertaining to peer review. Finally, we issue a detailed call for action for the scientific community, NLP and AI researchers, policymakers, and funding bodies to help bring the research in NLP for peer review forward. We hope that our work will help set the agenda for research in machine-assisted scientific quality control in the age of AI, within the NLP community and beyond.

Vaibhav Mishra, Somaditya Singh, Dhruv Ahlawat et al.

Materials discovery and development are critical for addressing global challenges. Yet, the exponential growth in materials science literature comprising vast amounts of textual data has created significant bottlenecks in knowledge extraction, synthesis, and scientific reasoning. Large Language Models (LLMs) offer unprecedented opportunities to accelerate materials research through automated analysis and prediction. Still, their effective deployment requires domain-specific adaptation for understanding and solving domain-relevant tasks. Here, we present LLaMat, a family of foundational models for materials science developed through continued pretraining of LLaMA models on an extensive corpus of materials literature and crystallographic data. Through systematic evaluation, we demonstrate that LLaMat excels in materials-specific NLP and structured information extraction while maintaining general linguistic capabilities. The specialized LLaMat-CIF variant demonstrates unprecedented capabilities in crystal structure generation, predicting stable crystals with high coverage across the periodic table. Intriguingly, despite LLaMA-3's superior performance in comparison to LLaMA-2, we observe that LLaMat-2 demonstrates unexpectedly enhanced domain-specific performance across diverse materials science tasks, including structured information extraction from text and tables, more particularly in crystal structure generation, a potential adaptation rigidity in overtrained LLMs. Altogether, the present work demonstrates the effectiveness of domain adaptation towards developing practically deployable LLM copilots for materials research. Beyond materials science, our findings reveal important considerations for domain adaptation of LLMs, such as model selection, training methodology, and domain-specific performance, which may influence the development of specialized scientific AI systems.

Chinmay Mittal, Krishna Kartik, Mausam et al.

Can the large language models (LLMs) solve challenging first-order combinatorial reasoning problems such as graph coloring, knapsack, and cryptarithmetic? By first-order, we mean these problems can be instantiated with potentially an infinite number of problem instances of varying sizes. They are also challenging being NP-hard and requiring several reasoning steps to reach a solution. While existing work has focused on coming up with datasets with hard benchmarks, there is limited work which exploits the first-order nature of the problem structure. To address this challenge, we present FCoReBench, a dataset of 40 such challenging problems, along with scripts to generate problem instances of varying sizes and automatically verify and generate their solutions. We first observe that LLMs, even when aided by symbolic solvers, perform rather poorly on our dataset, being unable to leverage the underlying structure of these problems. We specifically observe a drop in performance with increasing problem size. In response, we propose a new approach, SymPro-LM, which combines LLMs with both symbolic solvers and program interpreters, along with feedback from a few solved examples, to achieve huge performance gains. Our proposed approach is robust to changes in the problem size, and has the unique characteristic of not requiring any LLM call during inference time, unlike earlier approaches. As an additional experiment, we also demonstrate SymPro-LM's effectiveness on other logical reasoning benchmarks.

Prayushi Faldu, Indrajit Bhattacharya, Mausam

An essential requirement for a real-world Knowledge Base Question Answering (KBQA) system is the ability to detect the answerability of questions when generating logical forms. However, state-of-the-art KBQA models assume all questions to be answerable. Recent research has found that such models, when superficially adapted to detect answerability, struggle to satisfactorily identify the different categories of unanswerable questions, and simultaneously preserve good performance for answerable questions. Towards addressing this issue, we propose RetinaQA, a new KBQA model that unifies two key ideas in a single KBQA architecture: (a) discrimination over candidate logical forms, rather than generating these, for handling schema-related unanswerability, and (b) sketch-filling-based construction of candidate logical forms for handling data-related unaswerability. Our results show that RetinaQA significantly outperforms adaptations of state-of-the-art KBQA models in handling both answerable and unanswerable questions and demonstrates robustness across all categories of unanswerability. Notably, RetinaQA also sets a new state-of-the-art for answerable KBQA, surpassing existing models.

Vishal Vivek Saley, Rocktim Jyoti Das, Dinesh Raghu et al. · ibm-research

End-to-end Task-Oriented Dialog (TOD) systems typically require extensive training datasets to perform well. In contrast, large language model (LLM) based TOD systems can excel even with limited data due to their ability to learn tasks through in-context exemplars. However, these models lack alignment with the style of responses in training data and often generate comprehensive responses, making it difficult for users to grasp the information quickly. In response, we propose SyncTOD that synergizes LLMs with task-specific hints to improve alignment in low-data settings. SyncTOD employs small auxiliary models to provide hints and select exemplars for in-context prompts. With ChatGPT, SyncTOD achieves superior performance compared to LLM-based baselines and SoTA models in low-data settings, while retaining competitive performance in full-data settings.

Vishal Vivek Saley, Goonjan Saha, Rocktim Jyoti Das et al. · ibm-research

Medical task-oriented dialogue systems can assist doctors by collecting patient medical history, aiding in diagnosis, or guiding treatment selection, thereby reducing doctor burnout and expanding access to medical services. However, doctor-patient dialogue datasets are not readily available, primarily due to privacy regulations. Moreover, existing datasets lack comprehensive annotations involving medical slots and their different attributes, such as symptoms and their onset, progression, and severity. These comprehensive annotations are crucial for accurate diagnosis. Finally, most existing datasets are non-English, limiting their utility for the larger research community. In response, we introduce MediTOD, a new dataset of doctor-patient dialogues in English for the medical history-taking task. Collaborating with doctors, we devise a questionnaire-based labeling scheme tailored to the medical domain. Then, medical professionals create the dataset with high-quality comprehensive annotations, capturing medical slots and their attributes. We establish benchmarks in supervised and few-shot settings on MediTOD for natural language understanding, policy learning, and natural language generation subtasks, evaluating models from both TOD and biomedical domains. We make MediTOD publicly available for future research.

Vipul Rathore, Malik Hammad Faisal, Parag Singla et al.

Distantly Supervised Relation Extraction (DSRE) remains a long-standing challenge in NLP, where models must learn from noisy bag-level annotations while making sentence-level predictions. While existing state-of-the-art (SoTA) DSRE models rely on task-specific training, their integration with in-context learning (ICL) using large language models (LLMs) remains underexplored. A key challenge is that the LLM may not learn relation semantics correctly, due to noisy annotation. In response, we propose HYDRE -- HYbrid Distantly Supervised Relation Extraction framework. It first uses a trained DSRE model to identify the top-k candidate relations for a given test sentence, then uses a novel dynamic exemplar retrieval strategy that extracts reliable, sentence-level exemplars from training data, which are then provided in LLM prompt for outputting the final relation(s). We further extend HYDRE to cross-lingual settings for RE in low-resource languages. Using available English DSRE training data, we evaluate all methods on English as well as a newly curated benchmark covering four diverse low-resource Indic languages -- Oriya, Santali, Manipuri, and Tulu. HYDRE achieves up to 20 F1 point gains in English and, on average, 17 F1 points on Indic languages over prior SoTA DSRE models. Detailed ablations exhibit HYDRE's efficacy compared to other prompting strategies.

Mounica Maddela, Lingjue Xie, Daniel Preotiuc-Pietro et al.

Semantic parsing methods for converting text to SQL queries enable question answering over structured data and can greatly benefit analysts who routinely perform complex analytics on vast data stored in specialized relational databases. Although several benchmarks measure the abilities of text to SQL, the complexity of their questions is inherently limited by the level of expressiveness in query languages and none focus explicitly on questions involving complex analytical reasoning which require operations such as calculations over aggregate analytics, time series analysis or scenario understanding. In this paper, we introduce STARQA, the first public human-created dataset of complex analytical reasoning questions and answers on three specialized-domain databases. In addition to generating SQL directly using LLMs, we evaluate a novel approach (Text2SQLCode) that decomposes the task into a combination of SQL and Python: SQL is responsible for data fetching, and Python more naturally performs reasoning. Our results demonstrate that identifying and combining the abilities of SQL and Python is beneficial compared to using SQL alone, yet the dataset still remains quite challenging for the existing state-of-the-art LLMs.

Vipul Rathore, Aniruddha Deb, Ankish Chandresh et al.

Recently, very large language models (LLMs) have shown exceptional performance on several English NLP tasks with just in-context learning (ICL), but their utility in other languages is still underexplored. We investigate their effectiveness for NLP tasks in low-resource languages (LRLs), especially in the setting of zero-labelled cross-lingual transfer (0-CLT), where no labelled training data for the target language is available -- however training data from one or more related medium-resource languages (MRLs) is utilized, alongside the available unlabeled test data for a target language. We introduce Self-Supervised Prompting (SSP), a novel ICL approach tailored for the 0-CLT setting. SSP is based on the key observation that LLMs output more accurate labels if in-context exemplars are from the target language (even if their labels are slightly noisy). To operationalize this, since target language training data is not available in 0-CLT, SSP operates in two stages. In Stage I, using source MRL training data, target language's test data is noisily labeled. In Stage II, these noisy test data points are used as exemplars in ICL for further improved labelling. Additionally, our implementation of SSP uses a novel Integer Linear Programming (ILP)-based exemplar selection that balances similarity, prediction confidence (when available) and label coverage. Experiments on three tasks and eleven LRLs (from three regions) demonstrate that SSP strongly outperforms existing SOTA fine-tuned and prompting-based baselines in 0-CLT setup.

Riya Sawhney, Samrat Yadav, Indrajit Bhattacharya et al.

Real-world applications of KBQA require models to handle unanswerable questions with a limited volume of in-domain labeled training data. We propose the novel task of few-shot transfer for KBQA with unanswerable questions and contribute two new datasets for performance evaluation. We present FUn-FuSIC - a novel solution for our task that extends FuSIC KBQA, the state-of-the-art few-shot transfer model for answerable-only KBQA. We first note that FuSIC-KBQA's iterative repair makes a strong assumption that all questions are unanswerable. As a remedy, we propose Feedback for Unanswerability (FUn), which uses iterative repair using feedback from a suite of strong and weak verifiers, and an adaptation of self consistency for unanswerabilty to better assess the answerability of a question. Our experiments show that FUn-FuSIC significantly outperforms suitable adaptations of multiple LLM based and supervised SoTA models on our task, while establishing a new SoTA for answerable few-shot transfer as well.

D. Subhalingam, Keshav Kolluru, Mausam et al.

In the e-commerce domain, the accurate extraction of attribute-value pairs (e.g., Brand: Apple) from product titles and user search queries is crucial for enhancing search and recommendation systems. A major challenge with neural models for this task is the lack of high-quality training data, as the annotations for attribute-value pairs in the available datasets are often incomplete. To address this, we introduce GenToC, a model designed for training directly with partially-labeled data, eliminating the necessity for a fully annotated dataset. GenToC employs a marker-augmented generative model to identify potential attributes, followed by a token classification model that determines the associated values for each attribute. GenToC outperforms existing state-of-the-art models, exhibiting upto 56.3% increase in the number of accurate extractions. Furthermore, we utilize GenToC to regenerate the training dataset to expand attribute-value annotations. This bootstrapping substantially improves the data quality for training other standard NER models, which are typically faster but less capable in handling partially-labeled data, enabling them to achieve comparable performance to GenToC. Our results demonstrate GenToC's unique ability to learn from a limited set of partially-labeled data and improve the training of more efficient models, advancing the automated extraction of attribute-value pairs. Finally, our model has been successfully integrated into IndiaMART, India's largest B2B e-commerce platform, achieving a significant increase of 20.2% in the number of correctly identified attribute-value pairs over the existing deployed system while achieving a high precision of 89.5%.

Vishal Vivek Saley, Rocktim Jyoti Das, Dinesh Raghu et al.

Task-oriented dialog (TOD) agents often ground their responses on external knowledge bases (KBs). These KBs can be dynamic and may be updated frequently. Existing approaches for learning TOD agents assume the KB snapshot contemporary to each individual dialog is available during training. However, in real-world scenarios, only the latest KB snapshot is available during training and as a result, the train dialogs may contain facts conflicting with the latest KB. These dialog-KB inconsistencies in the training data may potentially confuse the TOD agent learning algorithm. In this work, we define the novel problem of learning a TOD agent with dialog-KB inconsistencies in the training data. We propose a Dialog-KB Arbitration Framework (DKAF) which reduces the dialog-KB inconsistencies by predicting the contemporary KB snapshot for each train dialog. These predicted KB snapshots are then used for training downstream TOD agents. As there are no existing datasets with dialog-KB inconsistencies, we systematically introduce inconsistencies in two publicly available dialog datasets. We show that TOD agents trained with DKAF perform better than existing baselines on both these datasets

Pranjal Aggarwal, Aman Madaan, Yiming Yang et al.

A popular approach for improving the correctness of output from large language models (LLMs) is Self-Consistency - poll the LLM multiple times and output the most frequent solution. Existing Self-Consistency techniques always generate a constant number of samples per question, where a better approach will be to non-uniformly distribute the available budget based on the amount of agreement in the samples generated so far. In response, we introduce Adaptive-Consistency, a cost-efficient, model-agnostic technique that dynamically adjusts the number of samples per question using a lightweight stopping criterion. Our experiments over 17 reasoning and code generation datasets and three LLMs demonstrate that Adaptive-Consistency reduces sample budget by up to 7.9 times with an average accuracy drop of less than 0.1%. Our code and data are available at https://www.sample-step-by-step.info

Ishaan Singh, Navdeep Kaur, Garima Gaur et al.

While Knowledge Graph Completion (KGC) on static facts is a matured field, Temporal Knowledge Graph Completion (TKGC), that incorporates validity time into static facts is still in its nascent stage. The KGC methods fall into multiple categories including embedding-based, rule-based, GNN-based, pretrained Language Model based approaches. However, such dimensions have not been explored in TKG. To that end, we propose a novel temporal neuro-symbolic model, NeuSTIP, that performs link prediction and time interval prediction in a TKG. NeuSTIP learns temporal rules in the presence of the Allen predicates that ensure the temporal consistency between neighboring predicates in a given rule. We further design a unique scoring function that evaluates the confidence of the candidate answers while performing link prediction and time interval prediction by utilizing the learned rules. Our empirical evaluation on two time interval based TKGC datasets suggests that our model outperforms state-of-the-art models for both link prediction and the time interval prediction task.

Kevin Leyton-Brown, Mausam, Yatin Nandwani et al.

Peer-reviewed conferences, the main publication venues in CS, rely critically on matching highly qualified reviewers for each paper. Because of the growing scale of these conferences, the tight timelines on which they operate, and a recent surge in explicitly dishonest behavior, there is now no alternative to performing this matching in an automated way. This paper studies a novel reviewer-paper matching approach that was recently deployed in the 35th AAAI Conference on Artificial Intelligence (AAAI 2021), and has since been adopted (wholly or partially) by other conferences including ICML 2022, AAAI 2022, and IJCAI 2022. This approach has three main elements: (1) collecting and processing input data to identify problematic matches and generate reviewer-paper scores; (2) formulating and solving an optimization problem to find good reviewer-paper matchings; and (3) a two-phase reviewing process that shifts reviewing resources away from papers likely to be rejected and towards papers closer to the decision boundary. This paper also describes an evaluation of these innovations based on an extensive post-hoc analysis on real data -- including a comparison with the matching algorithm used in AAAI's previous (2020) iteration -- and supplements this with additional numerical experimentation.

Yatin Nandwani, Vidit Jain, Mausam et al.

Recently many neural models have been proposed to solve combinatorial puzzles by implicitly learning underlying constraints using their solved instances, such as sudoku or graph coloring (GCP). One drawback of the proposed architectures, which are often based on Graph Neural Networks (GNN), is that they cannot generalize across the size of the output space from which variables are assigned a value, for example, set of colors in a GCP, or board-size in sudoku. We call the output space for the variables as 'value-set'. While many works have demonstrated generalization of GNNs across graph size, there has been no study on how to design a GNN for achieving value-set invariance for problems that come from the same domain. For example, learning to solve 16 x 16 sudoku after being trained on only 9 x 9 sudokus. In this work, we propose novel methods to extend GNN based architectures to achieve value-set invariance. Specifically, our model builds on recently proposed Recurrent Relational Networks. Our first approach exploits the graph-size invariance of GNNs by converting a multi-class node classification problem into a binary node classification problem. Our second approach works directly with multiple classes by adding multiple nodes corresponding to the values in the value-set, and then connecting variable nodes to value nodes depending on the problem initialization. Our experimental evaluation on three different combinatorial problems demonstrates that both our models perform well on our novel problem, compared to a generic neural reasoner. Between two of our models, we observe an inherent trade-off: while the binarized model gives better performance when trained on smaller value-sets, multi-valued model is much more memory efficient, resulting in improved performance when trained on larger value-sets, where binarized model fails to train.

Vipul Rathore, Kartikeya Badola, Mausam et al.

Neural models for distantly supervised relation extraction (DS-RE) encode each sentence in an entity-pair bag separately. These are then aggregated for bag-level relation prediction. Since, at encoding time, these approaches do not allow information to flow from other sentences in the bag, we believe that they do not utilize the available bag data to the fullest. In response, we explore a simple baseline approach (PARE) in which all sentences of a bag are concatenated into a passage of sentences, and encoded jointly using BERT. The contextual embeddings of tokens are aggregated using attention with the candidate relation as query -- this summary of whole passage predicts the candidate relation. We find that our simple baseline solution outperforms existing state-of-the-art DS-RE models in both monolingual and multilingual DS-RE datasets.

Tanishq Gupta, Mohd Zaki, N. M. Anoop Krishnan et al.

An overwhelmingly large amount of knowledge in the materials domain is generated and stored as text published in peer-reviewed scientific literature. Recent developments in natural language processing, such as bidirectional encoder representations from transformers (BERT) models, provide promising tools to extract information from these texts. However, direct application of these models in the materials domain may yield suboptimal results as the models themselves may not be trained on notations and jargon that are specific to the domain. Here, we present a materials-aware language model, namely, MatSciBERT, which is trained on a large corpus of scientific literature published in the materials domain. We further evaluate the performance of MatSciBERT on three downstream tasks, namely, abstract classification, named entity recognition, and relation extraction, on different materials datasets. We show that MatSciBERT outperforms SciBERT, a language model trained on science corpus, on all the tasks. Further, we discuss some of the applications of MatSciBERT in the materials domain for extracting information, which can, in turn, contribute to materials discovery or optimization. Finally, to make the work accessible to the larger materials community, we make the pretrained and finetuned weights and the models of MatSciBERT freely accessible.

Dinesh Raghu, Atishya Jain, Mausam et al.

End-to-End task-oriented dialogue systems generate responses based on dialog history and an accompanying knowledge base (KB). Inferring those KB entities that are most relevant for an utterance is crucial for response generation. Existing state of the art scales to large KBs by softly filtering over irrelevant KB information. In this paper, we propose a novel filtering technique that consists of (1) a pairwise similarity based filter that identifies relevant information by respecting the n-ary structure in a KB record. and, (2) an auxiliary loss that helps in separating contextually unrelated KB information. We also propose a new metric -- multiset entity F1 which fixes a correctness issue in the existing entity F1 metric. Experimental results on three publicly available task-oriented dialog datasets show that our proposed approach outperforms existing state-of-the-art models.

Dinesh Raghu, Shantanu Agarwal, Sachindra Joshi et al.

We propose a novel problem within end-to-end learning of task-oriented dialogs (TOD), in which the dialog system mimics a troubleshooting agent who helps a user by diagnosing their problem (e.g., car not starting). Such dialogs are grounded in domain-specific flowcharts, which the agent is supposed to follow during the conversation. Our task exposes novel technical challenges for neural TOD, such as grounding an utterance to the flowchart without explicit annotation, referring to additional manual pages when user asks a clarification question, and ability to follow unseen flowcharts at test time. We release a dataset (FloDial) consisting of 2,738 dialogs grounded on 12 different troubleshooting flowcharts. We also design a neural model, FloNet, which uses a retrieval-augmented generation architecture to train the dialog agent. Our experiments find that FloNet can do zero-shot transfer to unseen flowcharts, and sets a strong baseline for future research.

Prashant Pandey, Ajey Pai, Nisarg Bhatt et al.

Contrastive Learning (CL) is a recent representation learning approach, which encourages inter-class separability and intra-class compactness in learned image representations. Since medical images often contain multiple semantic classes in an image, using CL to learn representations of local features (as opposed to global) is important. In this work, we present a novel semi-supervised 2D medical segmentation solution that applies CL on image patches, instead of full images. These patches are meaningfully constructed using the semantic information of different classes obtained via pseudo labeling. We also propose a novel consistency regularization (CR) scheme, which works in synergy with CL. It addresses the problem of confirmation bias, and encourages better clustering in the feature space. We evaluate our method on four public medical segmentation datasets and a novel histopathology dataset that we introduce. Our method obtains consistent improvements over state-of-the-art semi-supervised segmentation approaches for all datasets.

Ananye Agarwal, Pradeep Shenoy, Mausam

Neural models and symbolic algorithms have recently been combined for tasks requiring both perception and reasoning. Neural models ground perceptual input into a conceptual vocabulary, on which a classical reasoning algorithm is applied to generate output. A key limitation is that such neural-to-symbolic models can only be trained end-to-end for tasks where the output space is symbolic. In this paper, we study neural-symbolic-neural models for reasoning tasks that require a conversion from an image input (e.g., a partially filled sudoku) to an image output (e.g., the image of the completed sudoku). While designing such a three-step hybrid architecture may be straightforward, the key technical challenge is end-to-end training -- how to backpropagate without intermediate supervision through the symbolic component. We propose NSNnet, an architecture that combines an image reconstruction loss with a novel output encoder to generate a supervisory signal, develops update algorithms that leverage policy gradient methods for supervision, and optimizes loss using a novel subsampling heuristic. We experiment on problem settings where symbolic algorithms are easily specified: a visual maze solving task and a visual Sudoku solver where the supervision is in image form. Experiments show high accuracy with significantly less data compared to purely neural approaches.

Shreshth Tuli, Rajas Bansal, Rohan Paul et al.

Robots assisting us in factories or homes must learn to make use of objects as tools to perform tasks, e.g., a tray for carrying objects. We consider the problem of learning commonsense knowledge of when a tool may be useful and how its use may be composed with other tools to accomplish a high-level task instructed by a human. We introduce a novel neural model, termed TANGO, for predicting task-specific tool interactions, trained using demonstrations from human teachers instructing a virtual robot. TANGO encodes the world state, comprising objects and symbolic relationships between them, using a graph neural network. The model learns to attend over the scene using knowledge of the goal and the action history, finally decoding the symbolic action to execute. Crucially, we address generalization to unseen environments where some known tools are missing, but alternative unseen tools are present. We show that by augmenting the representation of the environment with pre-trained embeddings derived from a knowledge-base, the model can generalize effectively to novel environments. Experimental results show a 60.5-78.9% absolute improvement over the baseline in predicting successful symbolic plans in unseen settings for a simulated mobile manipulator.

Harkanwar Singh, Prachi Jain, Mausam et al.

Knowledge Graph Completion (KGC) predicts missing facts in an incomplete Knowledge Graph. Almost all of existing KGC research is applicable to only one KG at a time, and in one language only. However, different language speakers may maintain separate KGs in their language and no individual KG is expected to be complete. Moreover, common entities or relations in these KGs have different surface forms and IDs, leading to ID proliferation. Entity alignment (EA) and relation alignment (RA) tasks resolve this by recognizing pairs of entity (relation) IDs in different KGs that represent the same entity (relation). This can further help prediction of missing facts, since knowledge from one KG is likely to benefit completion of another. High confidence predictions may also add valuable information for the alignment tasks. In response, we study the novel task of jointly training multilingual KGC, relation alignment and entity alignment models. We present ALIGNKGC, which uses some seed alignments to jointly optimize all three of KGC, EA and RA losses. A key component of ALIGNKGC is an embedding based soft notion of asymmetric overlap defined on the (subject, object) set signatures of relations this aids in better predicting relations that are equivalent to or implied by other relations. Extensive experiments with DBPedia in five languages establish the benefits of joint training for all tasks, achieving 10-32 MRR improvements of ALIGNKGC over a strong state-of-the-art single-KGC system completion model over each monolingual KG . Further, ALIGNKGC achieves reasonable gains in EA and RA tasks over a vanilla completion model over a KG that combines all facts without alignment, underscoring the value of joint training for these tasks.

Keshav Kolluru, Mayank Singh Chauhan, Yatin Nandwani et al.

Pre-trained language models (LMs) like BERT have shown to store factual knowledge about the world. This knowledge can be used to augment the information present in Knowledge Bases, which tend to be incomplete. However, prior attempts at using BERT for task of Knowledge Base Completion (KBC) resulted in performance worse than embedding based techniques that rely only on the graph structure. In this work we develop a novel model, Cross-Entity Aware Reranker (CEAR), that uses BERT to re-rank the output of existing KBC models with cross-entity attention. Unlike prior work that scores each entity independently, CEAR uses BERT to score the entities together, which is effective for exploiting its factual knowledge. CEAR achieves a new state of art for the OLPBench dataset.

Abhyuday Bhartiya, Kartikeya Badola, Mausam

Distant supervision (DS) is a well established technique for creating large-scale datasets for relation extraction (RE) without using human annotations. However, research in DS-RE has been mostly limited to the English language. Constraining RE to a single language inhibits utilization of large amounts of data in other languages which could allow extraction of more diverse facts. Very recently, a dataset for multilingual DS-RE has been released. However, our analysis reveals that the proposed dataset exhibits unrealistic characteristics such as 1) lack of sentences that do not express any relation, and 2) all sentences for a given entity pair expressing exactly one relation. We show that these characteristics lead to a gross overestimation of the model performance. In response, we propose a new dataset, DiS-ReX, which alleviates these issues. Our dataset has more than 1.5 million sentences, spanning across 4 languages with 36 relation classes + 1 no relation (NA) class. We also modify the widely used bag attention models by encoding sentences using mBERT and provide the first benchmark results on multilingual DS-RE. Unlike the competing dataset, we show that our dataset is challenging and leaves enough room for future research to take place in this field.

Keshav Kolluru, Vaibhav Adlakha, Samarth Aggarwal et al.

A recent state-of-the-art neural open information extraction (OpenIE) system generates extractions iteratively, requiring repeated encoding of partial outputs. This comes at a significant computational cost. On the other hand, sequence labeling approaches for OpenIE are much faster, but worse in extraction quality. In this paper, we bridge this trade-off by presenting an iterative labeling-based system that establishes a new state of the art for OpenIE, while extracting 10x faster. This is achieved through a novel Iterative Grid Labeling (IGL) architecture, which treats OpenIE as a 2-D grid labeling task. We improve its performance further by applying coverage (soft) constraints on the grid at training time. Moreover, on observing that the best OpenIE systems falter at handling coordination structures, our OpenIE system also incorporates a new coordination analyzer built with the same IGL architecture. This IGL based coordination analyzer helps our OpenIE system handle complicated coordination structures, while also establishing a new state of the art on the task of coordination analysis, with a 12.3 pts improvement in F1 over previous analyzers. Our OpenIE system, OpenIE6, beats the previous systems by as much as 4 pts in F1, while being much faster.

Danish Contractor, Shashank Goel, Mausam et al.

Our goal is to answer real-world tourism questions that seek Points-of-Interest (POI) recommendations. Such questions express various kinds of spatial and non-spatial constraints, necessitating a combination of textual and spatial reasoning. In response, we develop the first joint spatio-textual reasoning model, which combines geo-spatial knowledge with information in textual corpora to answer questions. We first develop a modular spatial-reasoning network that uses geo-coordinates of location names mentioned in a question, and of candidate answer POIs, to reason over only spatial constraints. We then combine our spatial-reasoner with a textual reasoner in a joint model and present experiments on a real world POI recommendation task. We report substantial improvements over existing models with-out joint spatio-textual reasoning.