Jordan Boyd-Graber

Chenglei Si, Zhe Gan, Zhengyuan Yang et al. · microsoft-research

Large language models (LLMs) show impressive abilities via few-shot prompting. Commercialized APIs such as OpenAI GPT-3 further increase their use in real-world language applications. However, the crucial problem of how to improve the reliability of GPT-3 is still under-explored. While reliability is a broad and vaguely defined term, we decompose reliability into four main facets that correspond to the existing framework of ML safety and are well-recognized to be important: generalizability, social biases, calibration, and factuality. Our core contribution is to establish simple and effective prompts that improve GPT-3's reliability as it: 1) generalizes out-of-distribution, 2) balances demographic distribution and uses natural language instructions to reduce social biases, 3) calibrates output probabilities, and 4) updates the LLM's factual knowledge and reasoning chains. With appropriate prompts, GPT-3 is more reliable than smaller-scale supervised models on all these facets. We release all processed datasets, evaluation scripts, and model predictions. Our systematic empirical study not only sheds new insights on the reliability of prompting LLMs, but more importantly, our prompting strategies can help practitioners more reliably use LLMs like GPT-3.

Fenfei Guo, Chen Zhang, Zhirui Zhang et al. · tencent-ai

This paper develops automatic song translation (AST) for tonal languages and addresses the unique challenge of aligning words' tones with melody of a song in addition to conveying the original meaning. We propose three criteria for effective AST -- preserving meaning, singability and intelligibility -- and design metrics for these criteria. We develop a new benchmark for English--Mandarin song translation and develop an unsupervised AST system, Guided AliGnment for Automatic Song Translation (GagaST), which combines pre-training with three decoding constraints. Both automatic and human evaluations show GagaST successfully balances semantics and singability.

Samuel Barham, Orion Weller, Michelle Yuan et al.

To foster the development of new models for collaborative AI-assisted report generation, we introduce MegaWika, consisting of 13 million Wikipedia articles in 50 diverse languages, along with their 71 million referenced source materials. We process this dataset for a myriad of applications, going beyond the initial Wikipedia citation extraction and web scraping of content, including translating non-English articles for cross-lingual applications and providing FrameNet parses for automated semantic analysis. MegaWika is the largest resource for sentence-level report generation and the only report generation dataset that is multilingual. We manually analyze the quality of this resource through a semantically stratified sample. Finally, we provide baseline results and trained models for crucial steps in automated report generation: cross-lingual question answering and citation retrieval.

Yoshinari Fujinuma, Jordan Boyd-Graber, Katharina Kann

Pretrained multilingual models enable zero-shot learning even for unseen languages, and that performance can be further improved via adaptation prior to finetuning. However, it is unclear how the number of pretraining languages influences a model's zero-shot learning for languages unseen during pretraining. To fill this gap, we ask the following research questions: (1) How does the number of pretraining languages influence zero-shot performance on unseen target languages? (2) Does the answer to that question change with model adaptation? (3) Do the findings for our first question change if the languages used for pretraining are all related? Our experiments on pretraining with related languages indicate that choosing a diverse set of languages is crucial. Without model adaptation, surprisingly, increasing the number of pretraining languages yields better results up to adding related languages, after which performance plateaus. In contrast, with model adaptation via continued pretraining, pretraining on a larger number of languages often gives further improvement, suggesting that model adaptation is crucial to exploit additional pretraining languages.

Chenglei Si, Chen Zhao, Sewon Min et al.

For users to trust model predictions, they need to understand model outputs, particularly their confidence - calibration aims to adjust (calibrate) models' confidence to match expected accuracy. We argue that the traditional calibration evaluation does not promote effective calibrations: for example, it can encourage always assigning a mediocre confidence score to all predictions, which does not help users distinguish correct predictions from wrong ones. Building on those observations, we propose a new calibration metric, MacroCE, that better captures whether the model assigns low confidence to wrong predictions and high confidence to correct predictions. Focusing on the practical application of open-domain question answering, we examine conventional calibration methods applied on the widely-used retriever-reader pipeline, all of which do not bring significant gains under our new MacroCE metric. Toward better calibration, we propose a new calibration method (ConsCal) that uses not just final model predictions but whether multiple model checkpoints make consistent predictions. Altogether, we provide an alternative view of calibration along with a new metric, re-evaluation of existing calibration methods on our metric, and proposal of a more effective calibration method.

Chenglei Si, Navita Goyal, Sherry Tongshuang Wu et al.

Large Language Models (LLMs) are increasingly used for accessing information on the web. Their truthfulness and factuality are thus of great interest. To help users make the right decisions about the information they get, LLMs should not only provide information but also help users fact-check it. Our experiments with 80 crowdworkers compare language models with search engines (information retrieval systems) at facilitating fact-checking. We prompt LLMs to validate a given claim and provide corresponding explanations. Users reading LLM explanations are significantly more efficient than those using search engines while achieving similar accuracy. However, they over-rely on the LLMs when the explanation is wrong. To reduce over-reliance on LLMs, we ask LLMs to provide contrastive information - explain both why the claim is true and false, and then we present both sides of the explanation to users. This contrastive explanation mitigates users' over-reliance on LLMs, but cannot significantly outperform search engines. Further, showing both search engine results and LLM explanations offers no complementary benefits compared to search engines alone. Taken together, our study highlights that natural language explanations by LLMs may not be a reliable replacement for reading the retrieved passages, especially in high-stakes settings where over-relying on wrong AI explanations could lead to critical consequences.

Neha Srikanth, Rupak Sarkar, Heran Mane et al.

Questions posed by information-seeking users often contain implicit false or potentially harmful assumptions. In a high-risk domain such as maternal and infant health, a question-answering system must recognize these pragmatic constraints and go beyond simply answering user questions, examining them in context to respond helpfully. To achieve this, we study assumptions and implications, or pragmatic inferences, made when mothers ask questions about pregnancy and infant care by collecting a dataset of 2,727 inferences from 500 questions across three diverse sources. We study how health experts naturally address these inferences when writing answers, and illustrate that informing existing QA pipelines with pragmatic inferences produces responses that are more complete, mitigating the propagation of harmful beliefs.

Maharshi Gor, Yoo Yeon Sung, Yu Hou et al.

AI systems are fallible, and humans can make mistakes in deciding whether to trust AI over their own judgment. Thus, improving human-AI collaboration requires understanding when, why, and how humans decide to rely on AI. We study two distinct reliance decisions: the delegation choice -- deciding when to let AI act autonomously without knowing its output, and the adoption choice -- evaluating AI suggestions and deciding how to use them. Both of these decoupled reliance patterns shape collaboration, but prior work rarely studies them together in realistic settings with the same users. We address this gap by studying collaborative human--AI teams competing in a question-answering game in which humans can choose when and how to work with AI agents to win. Our 24 matches pair 23 expert humans with 16 AI agents, capturing 387 delegation and 1440 adoption decisions. While human--AI collaboration performs better than either AI or humans alone, humans make suboptimal collaboration decisions, both under-relying on correct AI suggestions (3.9% of opportunities missed) and over-relying when AI misleads them (1.7%). Both parties contribute wrong answers: reported model confidence is near chance when humans and AI disagree, while confirmation bias drives higher under-reliance (64.5%) when an AI suggestion agrees with humans' initial incorrect answer. To close this gap, we recommend calibrated confidence, evidence-grounded explanations, and mechanisms that help users refine trust.

Sander Schulhoff, Jeremy Pinto, Anaum Khan et al.

Large Language Models (LLMs) are deployed in interactive contexts with direct user engagement, such as chatbots and writing assistants. These deployments are vulnerable to prompt injection and jailbreaking (collectively, prompt hacking), in which models are manipulated to ignore their original instructions and follow potentially malicious ones. Although widely acknowledged as a significant security threat, there is a dearth of large-scale resources and quantitative studies on prompt hacking. To address this lacuna, we launch a global prompt hacking competition, which allows for free-form human input attacks. We elicit 600K+ adversarial prompts against three state-of-the-art LLMs. We describe the dataset, which empirically verifies that current LLMs can indeed be manipulated via prompt hacking. We also present a comprehensive taxonomical ontology of the types of adversarial prompts.

Ishani Mondal, Zongxia Li, Yufang Hou et al.

Automating the creation of scientific diagrams from academic papers can significantly streamline the development of tutorials, presentations, and posters, thereby saving time and accelerating the process. Current text-to-image models struggle with generating accurate and visually appealing diagrams from long-context inputs. We propose SciDoc2Diagram, a task that extracts relevant information from scientific papers and generates diagrams, along with a benchmarking dataset, SciDoc2DiagramBench. We develop a multi-step pipeline SciDoc2Diagrammer that generates diagrams based on user intentions using intermediate code generation. We observed that initial diagram drafts were often incomplete or unfaithful to the source, leading us to develop SciDoc2Diagrammer-Multi-Aspect-Feedback (MAF), a refinement strategy that significantly enhances factual correctness and visual appeal and outperforms existing models on both automatic and human judgement.

Yoo Yeon Sung, Jordan Boyd-Graber, Naeemul Hassan

Polarization and the marketplace for impressions have conspired to make navigating information online difficult for users, and while there has been a significant effort to detect false or misleading text, multimodal datasets have received considerably less attention. To complement existing resources, we present multimodal Video Misleading Headline (VMH), a dataset that consists of videos and whether annotators believe the headline is representative of the video's contents. After collecting and annotating this dataset, we analyze multimodal baselines for detecting misleading headlines. Our annotation process also focuses on why annotators view a video as misleading, allowing us to better understand the interplay of annotators' background and the content of the videos.

Ishani Mondal, Yiwen Song, Mihir Parmar et al.

Long-form visual storytelling requires maintaining continuity across shots, including consistent characters, stable environments, and smooth scene transitions. While existing generative models can produce strong individual frames, they fail to preserve such continuity, leading to appearance changes, inconsistent backgrounds, and abrupt scene shifts. We introduce CANVAS (Continuity-Aware Narratives via Visual Agentic Storyboarding), a multi-agent framework that explicitly plans visual continuity in multi-shot narratives. CANVAS enforces coherence through character continuity, persistent background anchors, and location-aware scene planning for smooth transitions within the same setting We evaluate CANVAS on two storyboard generation benchmarks ST-BENCH and ViStoryBench and introduce a new challenging benchmark HardContinuityBench for long-range narrative consistency. CANVAS consistently outperforms the best-performing baseline, improving background continuity by 21.6%, character consistency by 9.6% and props consistency by 7.6%.

Neha Srikanth, Jordan Boyd-Graber, Rachel Rudinger

We introduce DiscoTrace, a method to identify the rhetorical strategies that answerers use when responding to information-seeking questions. DiscoTrace represents answers as a sequence of question-related discourse acts paired with interpretations of the original question, annotated on top of rhetorical structure theory parses. Applying DiscoTrace to answers from nine different human communities reveals that communities have diverse preferences for answer construction. In contrast, LLMs do not exhibit rhetorical diversity in their answers, even when prompted to mimic specific human community answering guidelines. LLMs also systematically opt for breadth, addressing interpretations of questions that human answerers choose not to address. Our findings can guide the development of pragmatic LLM answerers that consider a range of strategies informed by context in QA.

Zheyuan Zhang, Runze Li, Tasnim Kabir et al.

Image geo-localization is the task of predicting the specific location of an image and requires complex reasoning across visual, geographical, and cultural contexts. While prior Vision Language Models (VLMs) have the best accuracy at this task, there is a dearth of high-quality datasets and models for analytical reasoning. We first create NaviClues, a high-quality dataset derived from GeoGuessr, a popular geography game, to supply examples of expert reasoning from language. Using this dataset, we present Navig, a comprehensive image geo-localization framework integrating global and fine-grained image information. By reasoning with language, Navig reduces the average distance error by 14% compared to previous state-of-the-art models while requiring fewer than 1000 training samples. Our dataset and code are available at https://github.com/SparrowZheyuan18/Navig/.

Saptarashmi Bandyopadhyay, Shraman Pal, Hao Zou et al.

Question Answering (QA) systems require a large amount of annotated data which is costly and time-consuming to gather. Converting datasets of existing QA benchmarks are challenging due to different formats and complexities. To address these issues, we propose an algorithm to automatically generate shorter questions resembling day-to-day human communication in the Natural Questions (NQ) dataset from longer trivia questions in Quizbowl (QB) dataset by leveraging conversion in style among the datasets. This provides an automated way to generate more data for our QA systems. To ensure quality as well as quantity of data, we detect and remove ill-formed questions using a neural classifier. We demonstrate that in a low resource setting, using the generated data improves the QA performance over the baseline system on both NQ and QB data. Our algorithm improves the scalability of training data while maintaining quality of data for QA systems.

Kyle Seelman, Mozhi Zhang, Jordan Boyd-Graber

Topic models are valuable for understanding extensive document collections, but they don't always identify the most relevant topics. Classical probabilistic and anchor-based topic models offer interactive versions that allow users to guide the models towards more pertinent topics. However, such interactive features have been lacking in neural topic models. To correct this lacuna, we introduce a user-friendly interaction for neural topic models. This interaction permits users to assign a word label to a topic, leading to an update in the topic model where the words in the topic become closely aligned with the given label. Our approach encompasses two distinct kinds of neural topic models. The first includes models where topic embeddings are trainable and evolve during the training process. The second kind involves models where topic embeddings are integrated post-training, offering a different approach to topic refinement. To facilitate user interaction with these neural topic models, we have developed an interactive interface. This interface enables users to engage with and re-label topics as desired. We evaluate our method through a human study, where users can relabel topics to find relevant documents. Using our method, user labeling improves document rank scores, helping to find more relevant documents to a given query when compared to no user labeling.

Wanrong He, Andrew Mao, Jordan Boyd-Graber

For humans and computers, the first step in answering an open-domain question is retrieving a set of relevant documents from a large corpus. However, the strategies that computers use fundamentally differ from those of humans. To better understand these differences, we design a gamified interface for data collection -- Cheater's Bowl -- where a human answers complex questions with access to both traditional and modern search tools. We collect a dataset of human search sessions, analyze human search strategies, and compare them to state-of-the-art multi-hop QA models. Humans query logically, apply dynamic search chains, and use world knowledge to boost searching. We demonstrate how human queries can improve the accuracy of existing systems and propose improving the future design of QA models.

Alexander Hoyle, Lorena Calvo-Bartolomé, Jordan Boyd-Graber et al.

Topic model and document-clustering evaluations either use automated metrics that align poorly with human preferences or require expert labels that are intractable to scale. We design a scalable human evaluation protocol and a corresponding automated approximation that reflect practitioners' real-world usage of models. Annotators -- or an LLM-based proxy -- review text items assigned to a topic or cluster, infer a category for the group, then apply that category to other documents. Using this protocol, we collect extensive crowdworker annotations of outputs from a diverse set of topic models on two datasets. We then use these annotations to validate automated proxies, finding that the best LLM proxies are statistically indistinguishable from a human annotator and can therefore serve as a reasonable substitute in automated evaluations. Package, web interface, and data are at https://github.com/ahoho/proxann

Chen Zhao, Chenyan Xiong, Jordan Boyd-Graber et al.

Complex question answering often requires finding a reasoning chain that consists of multiple evidence pieces. Current approaches incorporate the strengths of structured knowledge and unstructured text, assuming text corpora is semi-structured. Building on dense retrieval methods, we propose a new multi-step retrieval approach (BeamDR) that iteratively forms an evidence chain through beam search in dense representations. When evaluated on multi-hop question answering, BeamDR is competitive to state-of-the-art systems, without using any semi-structured information. Through query composition in dense space, BeamDR captures the implicit relationships between evidence in the reasoning chain. The code is available at https://github.com/ henryzhao5852/BeamDR.

Julian Martin Eisenschlos, Bhuwan Dhingra, Jannis Bulian et al.

We release FoolMeTwice (FM2 for short), a large dataset of challenging entailment pairs collected through a fun multi-player game. Gamification encourages adversarial examples, drastically lowering the number of examples that can be solved using "shortcuts" compared to other popular entailment datasets. Players are presented with two tasks. The first task asks the player to write a plausible claim based on the evidence from a Wikipedia page. The second one shows two plausible claims written by other players, one of which is false, and the goal is to identify it before the time runs out. Players "pay" to see clues retrieved from the evidence pool: the more evidence the player needs, the harder the claim. Game-play between motivated players leads to diverse strategies for crafting claims, such as temporal inference and diverting to unrelated evidence, and results in higher quality data for the entailment and evidence retrieval tasks. We open source the dataset and the game code.

Zongxia Li, Wenhao Yu, Chengsong Huang et al.

Vision-Language Models (VLMs) often suffer from visual hallucinations, saying things that are not actually in the image, and language shortcuts, where they skip the visual part and just rely on text priors. These issues arise because most post-training methods for VLMs rely on simple verifiable answer matching and supervise only final outputs, leaving intermediate visual reasoning without explicit guidance. As a result, VLMs receive sparse visual signals and often learn to prioritize language-based reasoning over visual perception. To mitigate this, some existing methods add visual supervision using human annotations or distilled labels from external large models. However, human annotations are labor-intensive and costly, and because external signals cannot adapt to the evolving policy, they cause distributional shifts that can lead to reward hacking. In this paper, we introduce Vision-SR1, a self-rewarding method that improves visual reasoning without relying on external visual supervisions via reinforcement learning. Vision-SR1 decomposes VLM reasoning into two stages: visual perception and language reasoning. The model is first prompted to produce self-contained visual perceptions that are sufficient to answer the question without referring back the input image. To validate this self-containment, the same VLM model is then re-prompted to perform language reasoning using only the generated perception as input to compute reward. This self-reward is combined with supervision on final outputs, providing a balanced training signal that strengthens both visual perception and language reasoning. Our experiments demonstrate that Vision-SR1 improves visual reasoning, mitigates visual hallucinations, and reduces reliance on language shortcuts across diverse vision-language tasks.

Zongxia Li, Andrew Mao, Daniel Stephens et al.

Topic models are a popular tool for understanding text collections, but their evaluation has been a point of contention. Automated evaluation metrics such as coherence are often used, however, their validity has been questioned for neural topic models (NTMs) and can overlook a models benefits in real world applications. To this end, we conduct the first evaluation of neural, supervised and classical topic models in an interactive task based setting. We combine topic models with a classifier and test their ability to help humans conduct content analysis and document annotation. From simulated, real user and expert pilot studies, the Contextual Neural Topic Model does the best on cluster evaluation metrics and human evaluations; however, LDA is competitive with two other NTMs under our simulated experiment and user study results, contrary to what coherence scores suggest. We show that current automated metrics do not provide a complete picture of topic modeling capabilities, but the right choice of NTMs can be better than classical models on practical task.

Nishant Balepur, Feng Gu, Abhilasha Ravichander et al. · cmu

Question answering (QA), giving correct answers to questions, is a popular task, but we test reverse question answering (RQA): for an input answer, give a question with that answer. Past work tests QA and RQA separately, but we test them jointly, comparing their difficulty, aiding benchmark design, and checking reasoning consistency. We run 16 LLMs on QA and RQA with trivia questions/answers, revealing: 1) Versus QA, LLMs are much less accurate in RQA for numerical answers, but slightly more accurate in RQA for textual answers; 2) LLMs often answer their own invalid questions from RQA accurately in QA, so RQA errors are not from knowledge gaps alone; 3) RQA errors correlate with question difficulty and inversely correlate with answer frequencies in the Dolma corpus; and 4) LLMs struggle to provide valid multi-hop questions. By finding question and answer types that lead to RQA errors, we suggest improvements for LLM reasoning.

Matthew Shu, Nishant Balepur, Shi Feng et al.

Flashcard schedulers rely on 1) student models to predict the flashcards a student knows; and 2) teaching policies to pick which cards to show next via these predictions. Prior student models, however, just use study data like the student's past responses, ignoring the text on cards. We propose content-aware scheduling, the first schedulers exploiting flashcard content. To give the first evidence that such schedulers enhance student learning, we build KARL, a simple but effective content-aware student model employing deep knowledge tracing (DKT), retrieval, and BERT to predict student recall. We train KARL by collecting a new dataset of 123,143 study logs on diverse trivia questions. KARL bests existing student models in AUC and calibration error. To ensure our improved predictions lead to better student learning, we create a novel delta-based teaching policy to deploy KARL online. Based on 32 study paths from 27 users, KARL improves learning efficiency over SOTA, showing KARL's strength and encouraging researchers to look beyond historical study data to fully capture student abilities.

Zongxia Li, Lorena Calvo-Bartolomé, Alexander Hoyle et al.

A common use of NLP is to facilitate the understanding of large document collections, with a shift from using traditional topic models to Large Language Models. Yet the effectiveness of using LLM for large corpus understanding in real-world applications remains under-explored. This study measures the knowledge users acquire with unsupervised, supervised LLM-based exploratory approaches or traditional topic models on two datasets. While LLM-based methods generate more human-readable topics and show higher average win probabilities than traditional models for data exploration, they produce overly generic topics for domain-specific datasets that do not easily allow users to learn much about the documents. Adding human supervision to the LLM generation process improves data exploration by mitigating hallucination and over-genericity but requires greater human effort. In contrast, traditional. models like Latent Dirichlet Allocation (LDA) remain effective for exploration but are less user-friendly. We show that LLMs struggle to describe the haystack of large corpora without human help, particularly domain-specific data, and face scaling and hallucination limitations due to context length constraints.

Nishant Balepur, Alexa Siu, Nedim Lipka et al.

Query-focused summarization (QFS) gives a summary of documents to answer a query. Past QFS work assumes queries have one answer, ignoring debatable ones (Is law school worth it?). We introduce Debatable QFS (DQFS), a task to create summaries that answer debatable queries via documents with opposing perspectives; summaries must comprehensively cover all sources and balance perspectives, favoring no side. These goals elude LLM QFS systems, which: 1) lack structured content plans, failing to guide LLMs to write balanced summaries, and 2) use the same query to retrieve contexts across documents, failing to cover all perspectives specific to each document's content. To overcome this, we design MODS, a multi-LLM framework mirroring human panel discussions. MODS treats documents as individual Speaker LLMs and has a Moderator LLM that picks speakers to respond to tailored queries for planned topics. Speakers use tailored queries to retrieve relevant contexts from their documents and supply perspectives, which are tracked in a rich outline, yielding a content plan to guide the final summary. Experiments on ConflictingQA with controversial web queries and DebateQFS, our new dataset of debate queries from Debatepedia, show MODS beats SOTA by 38-59% in topic paragraph coverage and balance, based on new citation metrics. Users also find MODS's summaries to be readable and more balanced.

Feng Gu, Wichayaporn Wongkamjan, Jonathan K. Kummerfeld et al.

AIs can beat humans in game environments; however, how helpful those agents are to human remains understudied. We augment CICERO, a natural language agent that demonstrates superhuman performance in Diplomacy, to generate both move and message advice based on player intentions. A dozen Diplomacy games with novice and experienced players, with varying advice settings, show that some of the generated advice is beneficial. It helps novices compete with experienced players and in some instances even surpass them. The mere presence of advice can be advantageous, even if players do not follow it.

Lorena Calvo-Bartolomé, Valérie Aldana, Karla Cantarero et al.

Multilingual question answering (QA) systems must ensure factual consistency across languages, especially for objective queries such as What is jaundice?, while also accounting for cultural variation in subjective responses. We propose MIND, a user-in-the-loop fact-checking pipeline to detect factual and cultural discrepancies in multilingual QA knowledge bases. MIND highlights divergent answers to culturally sensitive questions (e.g., Who assists in childbirth?) that vary by region and context. We evaluate MIND on a bilingual QA system in the maternal and infant health domain and release a dataset of bilingual questions annotated for factual and cultural inconsistencies. We further test MIND on datasets from other domains to assess generalization. In all cases, MIND reliably identifies inconsistencies, supporting the development of more culturally aware and factually consistent QA systems.

Nishant Balepur, Matthew Shu, Alexander Hoyle et al.

Keyword mnemonics are memorable explanations that link new terms to simpler keywords. Prior work generates mnemonics for students, but they do not train models using mnemonics students prefer and aid learning. We build SMART, a mnemonic generator trained on feedback from real students learning new terms. To train SMART, we first fine-tune LLaMA-2 on a curated set of user-written mnemonics. We then use LLM alignment to enhance SMART: we deploy mnemonics generated by SMART in a flashcard app to find preferences on mnemonics students favor. We gather 2684 preferences from 45 students across two types: expressed (inferred from ratings) and observed (inferred from student learning), yielding three key findings. First, expressed and observed preferences disagree; what students think is helpful does not always capture what is truly helpful. Second, Bayesian models can synthesize complementary data from multiple preference types into a single effectiveness signal. SMART is tuned via Direct Preference Optimization on this signal, which resolves ties and missing labels in the typical method of pairwise comparisons, augmenting data for LLM output quality gains. Third, mnemonic experts assess SMART as matching GPT-4 at much lower deployment costs, showing the utility of capturing diverse student feedback to align LLMs in education.

Zongxia Li, Ishani Mondal, Yijun Liang et al.

Question answering (QA) can only make progress if we know if an answer is correct, but for many of the most challenging and interesting QA examples, current evaluation metrics to determine answer equivalence (AE) often do not align with human judgments, particularly more verbose, free-form answers from large language models (LLM). There are two challenges: a lack of data and that models are too big: LLM-based scorers can correlate better with human judges, but this task has only been tested on limited QA datasets, and even when available, update of the model is limited because LLMs are large and often expensive. We rectify both of these issues by providing clear and consistent guidelines for evaluating AE in machine QA adopted from professional human QA contests. We also introduce a combination of standard evaluation and a more efficient, robust, and lightweight discriminate AE classifier-based matching method (CFMatch, smaller than 1 MB), trained and validated to more accurately evaluate answer correctness in accordance with adopted expert AE rules that are more aligned with human judgments.

Yoo Yeon Sung, Ishani Mondal, Jordan Boyd-Graber

Dynamic adversarial question generation, where humans write examples to stump a model, aims to create examples that are realistic and informative. However, the advent of large language models (LLMs) has been a double-edged sword for human authors: more people are interested in seeing and pushing the limits of these models, but because the models are so much stronger an opponent, they are harder to defeat. To understand how these models impact adversarial question writing process, we enrich the writing guidance with LLMs and retrieval models for the authors to reason why their questions are not adversarial. While authors could create interesting, challenging adversarial questions, they sometimes resort to tricks that result in poor questions that are ambiguous, subjective, or confusing not just to a computer but also to humans. To address these issues, we propose new metrics and incentives for eliciting good, challenging questions and present a new dataset of adversarially authored questions.

Ishani Mondal, Michelle Yuan, Anandhavelu N et al.

Learning template based information extraction from documents is a crucial yet difficult task. Prior template-based IE approaches assume foreknowledge of the domain templates; however, real-world IE do not have pre-defined schemas and it is a figure-out-as you go phenomena. To quickly bootstrap templates in a real-world setting, we need to induce template slots from documents with zero or minimal supervision. Since the purpose of question answering intersect with the goal of information extraction, we use automatic question generation to induce template slots from the documents and investigate how a tiny amount of a proxy human-supervision on-the-fly (termed as InteractiveIE) can further boost the performance. Extensive experiments on biomedical and legal documents, where obtaining training data is expensive, reveal encouraging trends of performance improvement using InteractiveIE over AI-only baseline.

Chenglei Si, Weijia Shi, Chen Zhao et al.

While recent large language models (LLMs) improve on various question answering (QA) datasets, it remains difficult for a single model to generalize across question types that require distinct reasoning abilities. We provide empirical evidence that state-of-the-art LLMs suffer from poor generalizability on reasoning types beyond those seen in the prompt. To remedy this, we propose a Mixture-of-Reasoning-Experts (MoRE) framework that ensembles diverse specialized language models. We specialize the backbone language model with prompts optimized for different reasoning categories, including factual, multihop, mathematical, and commonsense reasoning. Our key insight is to leverage agreement among the specialized experts to select the best answer for each question, or to abstain from answering. This gives MoRE higher accuracy than any single specialized model on a collection of 12 QA datasets from four reasoning types. Beyond generalizability, the interpretable design of MoRE improves selective question answering results compared to baselines without incorporating inter-expert agreement. This framework is also more interpretable and useful to human consumers of QA outputs. Our human study confirms that presenting expert predictions and the answer selection process helps annotators more accurately calibrate when to trust the system's output. We release all code and data to facilitate future work.

Chen Zhao, Chenyan Xiong, Jordan Boyd-Graber et al.

Open-domain question answering answers a question based on evidence retrieved from a large corpus. State-of-the-art neural approaches require intermediate evidence annotations for training. However, such intermediate annotations are expensive, and methods that rely on them cannot transfer to the more common setting, where only question-answer pairs are available. This paper investigates whether models can learn to find evidence from a large corpus, with only distant supervision from answer labels for model training, thereby generating no additional annotation cost. We introduce a novel approach (DistDR) that iteratively improves over a weak retriever by alternately finding evidence from the up-to-date model and encouraging the model to learn the most likely evidence. Without using any evidence labels, DistDR is on par with fully-supervised state-of-the-art methods on both multi-hop and single-hop QA benchmarks. Our analysis confirms that DistDR finds more accurate evidence over iterations, which leads to model improvements.

Chenglei Si, Chen Zhao, Jordan Boyd-Graber

A flaw in QA evaluation is that annotations often only provide one gold answer. Thus, model predictions semantically equivalent to the answer but superficially different are considered incorrect. This work explores mining alias entities from knowledge bases and using them as additional gold answers (i.e., equivalent answers). We incorporate answers for two settings: evaluation with additional answers and model training with equivalent answers. We analyse three QA benchmarks: Natural Questions, TriviaQA, and SQuAD. Answer expansion increases the exact match score on all datasets for evaluation, while incorporating it helps model training over real-world datasets. We ensure the additional answers are valid through a human post hoc evaluation.

Peter Jansen, Jordan Boyd-Graber

Tamarian, a fictional language introduced in the Star Trek episode Darmok, communicates meaning through utterances of metaphorical references, such as "Darmok and Jalad at Tanagra" instead of "We should work together." This work assembles a Tamarian-English dictionary of utterances from the original episode and several follow-on novels, and uses this to construct a parallel corpus of 456 English-Tamarian utterances. A machine translation system based on a large language model (T5) is trained using this parallel corpus, and is shown to produce an accuracy of 76% when translating from English to Tamarian on known utterances.

Alexander Hoyle, Pranav Goel, Denis Peskov et al.

Topic model evaluation, like evaluation of other unsupervised methods, can be contentious. However, the field has coalesced around automated estimates of topic coherence, which rely on the frequency of word co-occurrences in a reference corpus. Contemporary neural topic models surpass classical ones according to these metrics. At the same time, topic model evaluation suffers from a validation gap: automated coherence, developed for classical models, has not been validated using human experimentation for neural models. In addition, a meta-analysis of topic modeling literature reveals a substantial standardization gap in automated topic modeling benchmarks. To address the validation gap, we compare automated coherence with the two most widely accepted human judgment tasks: topic rating and word intrusion. To address the standardization gap, we systematically evaluate a dominant classical model and two state-of-the-art neural models on two commonly used datasets. Automated evaluations declare a winning model when corresponding human evaluations do not, calling into question the validity of fully automatic evaluations independent of human judgments.

Michelle Yuan, Patrick Xia, Chandler May et al.

Neural coreference resolution models trained on one dataset may not transfer to new, low-resource domains. Active learning mitigates this problem by sampling a small subset of data for annotators to label. While active learning is well-defined for classification tasks, its application to coreference resolution is neither well-defined nor fully understood. This paper explores how to actively label coreference, examining sources of model uncertainty and document reading costs. We compare uncertainty sampling strategies and their advantages through thorough error analysis. In both synthetic and human experiments, labeling spans within the same document is more effective than annotating spans across documents. The findings contribute to a more realistic development of coreference resolution models.

Maharshi Gor, Kellie Webster, Jordan Boyd-Graber

The goal of question answering (QA) is to answer any question. However, major QA datasets have skewed distributions over gender, profession, and nationality. Despite that skew, model accuracy analysis reveals little evidence that accuracy is lower for people based on gender or nationality; instead, there is more variation on professions (question topic). But QA's lack of representation could itself hide evidence of bias, necessitating QA datasets that better represent global diversity.

Chen Zhao, Chenyan Xiong, Xin Qian et al.

We introduce DELFT, a factoid question answering system which combines the nuance and depth of knowledge graph question answering approaches with the broader coverage of free-text. DELFT builds a free-text knowledge graph from Wikipedia, with entities as nodes and sentences in which entities co-occur as edges. For each question, DELFT finds the subgraph linking question entity nodes to candidates using text sentences as edges, creating a dense and high coverage semantic graph. A novel graph neural network reasons over the free-text graph-combining evidence on the nodes via information along edge sentences-to select a final answer. Experiments on three question answering datasets show DELFT can answer entity-rich questions better than machine reading based models, bert-based answer ranking and memory networks. DELFT's advantage comes from both the high coverage of its free-text knowledge graph-more than double that of dbpedia relations-and the novel graph neural network which reasons on the rich but noisy free-text evidence.

Sewon Min, Jordan Boyd-Graber, Chris Alberti et al.

We review the EfficientQA competition from NeurIPS 2020. The competition focused on open-domain question answering (QA), where systems take natural language questions as input and return natural language answers. The aim of the competition was to build systems that can predict correct answers while also satisfying strict on-disk memory budgets. These memory budgets were designed to encourage contestants to explore the trade-off between storing retrieval corpora or the parameters of learned models. In this report, we describe the motivation and organization of the competition, review the best submissions, and analyze system predictions to inform a discussion of evaluation for open-domain QA.

Thomas Diggelmann, Jordan Boyd-Graber, Jannis Bulian et al.

We introduce CLIMATE-FEVER, a new publicly available dataset for verification of climate change-related claims. By providing a dataset for the research community, we aim to facilitate and encourage work on improving algorithms for retrieving evidential support for climate-specific claims, addressing the underlying language understanding challenges, and ultimately help alleviate the impact of misinformation on climate change. We adapt the methodology of FEVER [1], the largest dataset of artificially designed claims, to real-life claims collected from the Internet. While during this process, we could rely on the expertise of renowned climate scientists, it turned out to be no easy task. We discuss the surprising, subtle complexity of modeling real-world climate-related claims within the \textsc{fever} framework, which we believe provides a valuable challenge for general natural language understanding. We hope that our work will mark the beginning of a new exciting long-term joint effort by the climate science and AI community.

Francesco S. Varini, Jordan Boyd-Graber, Massimiliano Ciaramita et al.

Climate change communication in the mass media and other textual sources may affect and shape public perception. Extracting climate change information from these sources is an important task, e.g., for filtering content and e-discovery, sentiment analysis, automatic summarization, question-answering, and fact-checking. However, automating this process is a challenge, as climate change is a complex, fast-moving, and often ambiguous topic with scarce resources for popular text-based AI tasks. In this paper, we introduce \textsc{ClimaText}, a dataset for sentence-based climate change topic detection, which we make publicly available. We explore different approaches to identify the climate change topic in various text sources. We find that popular keyword-based models are not adequate for such a complex and evolving task. Context-based algorithms like BERT \cite{devlin2018bert} can detect, in addition to many trivial cases, a variety of complex and implicit topic patterns. Nevertheless, our analysis reveals a great potential for improvement in several directions, such as, e.g., capturing the discussion on indirect effects of climate change. Hence, we hope this work can serve as a good starting point for further research on this topic.

Tianze Shi, Chen Zhao, Jordan Boyd-Graber et al.

Large-scale semantic parsing datasets annotated with logical forms have enabled major advances in supervised approaches. But can richer supervision help even more? To explore the utility of fine-grained, lexical-level supervision, we introduce Squall, a dataset that enriches 11,276 WikiTableQuestions English-language questions with manually created SQL equivalents plus alignments between SQL and question fragments. Our annotation enables new training possibilities for encoder-decoder models, including approaches from machine translation previously precluded by the absence of alignments. We propose and test two methods: (1) supervised attention; (2) adopting an auxiliary objective of disambiguating references in the input queries to table columns. In 5-fold cross validation, these strategies improve over strong baselines by 4.4% execution accuracy. Oracle experiments suggest that annotated alignments can support further accuracy gains of up to 23.9%.

Michelle Yuan, Hsuan-Tien Lin, Jordan Boyd-Graber

Active learning strives to reduce annotation costs by choosing the most critical examples to label. Typically, the active learning strategy is contingent on the classification model. For instance, uncertainty sampling depends on poorly calibrated model confidence scores. In the cold-start setting, active learning is impractical because of model instability and data scarcity. Fortunately, modern NLP provides an additional source of information: pre-trained language models. The pre-training loss can find examples that surprise the model and should be labeled for efficient fine-tuning. Therefore, we treat the language modeling loss as a proxy for classification uncertainty. With BERT, we develop a simple strategy based on the masked language modeling loss that minimizes labeling costs for text classification. Compared to other baselines, our approach reaches higher accuracy within less sampling iterations and computation time.

Mozhi Zhang, Yoshinari Fujinuma, Michael J. Paul et al.

Cross-lingual word embeddings (CLWE) are often evaluated on bilingual lexicon induction (BLI). Recent CLWE methods use linear projections, which underfit the training dictionary, to generalize on BLI. However, underfitting can hinder generalization to other downstream tasks that rely on words from the training dictionary. We address this limitation by retrofitting CLWE to the training dictionary, which pulls training translation pairs closer in the embedding space and overfits the training dictionary. This simple post-processing step often improves accuracy on two downstream tasks, despite lowering BLI test accuracy. We also retrofit to both the training dictionary and a synthetic dictionary induced from CLWE, which sometimes generalizes even better on downstream tasks. Our results confirm the importance of fully exploiting training dictionary in downstream tasks and explains why BLI is a flawed CLWE evaluation.

Benjamin Borschinger, Jordan Boyd-Graber, Christian Buck et al.

We investigate a framework for machine reading, inspired by real world information-seeking problems, where a meta question answering system interacts with a black box environment. The environment encapsulates a competitive machine reader based on BERT, providing candidate answers to questions, and possibly some context. To validate the realism of our formulation, we ask humans to play the role of a meta-answerer. With just a small snippet of text around an answer, humans can outperform the machine reader, improving recall. Similarly, a simple machine meta-answerer outperforms the environment, improving both precision and recall on the Natural Questions dataset. The system relies on joint training of answer scoring and the selection of conditioning information.

Michelle Yuan, Mozhi Zhang, Benjamin Van Durme et al.

Cross-lingual word embeddings transfer knowledge between languages: models trained on high-resource languages can predict in low-resource languages. We introduce CLIME, an interactive system to quickly refine cross-lingual word embeddings for a given classification problem. First, CLIME ranks words by their salience to the downstream task. Then, users mark similarity between keywords and their nearest neighbors in the embedding space. Finally, CLIME updates the embeddings using the annotations. We evaluate CLIME on identifying health-related text in four low-resource languages: Ilocano, Sinhalese, Tigrinya, and Uyghur. Embeddings refined by CLIME capture more nuanced word semantics and have higher test accuracy than the original embeddings. CLIME often improves accuracy faster than an active learning baseline and can be easily combined with active learning to improve results.

Jordan Boyd-Graber, Benjamin Börschinger

In addition to the traditional task of getting machines to answer questions, a major research question in question answering is to create interesting, challenging questions that can help systems learn how to answer questions and also reveal which systems are the best at answering questions. We argue that creating a question answering dataset -- and the ubiquitous leaderboard that goes with it -- closely resembles running a trivia tournament: you write questions, have agents (either humans or machines) answer the questions, and declare a winner. However, the research community has ignored the decades of hard-learned lessons from decades of the trivia community creating vibrant, fair, and effective question answering competitions. After detailing problems with existing QA datasets, we outline the key lessons -- removing ambiguity, discriminating skill, and adjudicating disputes -- that can transfer to QA research and how they might be implemented for the QA community.

Denis Peskov, Joe Barrow, Pedro Rodriguez et al.

Natural language processing systems are often downstream of unreliable inputs: machine translation, optical character recognition, or speech recognition. For instance, virtual assistants can only answer your questions after understanding your speech. We investigate and mitigate the effects of noise from Automatic Speech Recognition systems on two factoid Question Answering (QA) tasks. Integrating confidences into the model and forced decoding of unknown words are empirically shown to improve the accuracy of downstream neural QA systems. We create and train models on a synthetic corpus of over 500,000 noisy sentences and evaluate on two human corpora from Quizbowl and Jeopardy! competitions.