Claims made by individuals or entities are oftentimes nuanced and cannot be clearly labeled as entirely "true" or "false" -- as is frequently the case with scientific and political claims. However, a claim (e.g., "vaccine A is better than vaccine B") can be dissected into its integral aspects and sub-aspects (e.g., efficacy, safety, distribution), which are individually easier to validate. This enables a more comprehensive, structured response that provides a well-rounded perspective on a given problem while also allowing the reader to prioritize specific angles of interest within the claim (e.g., safety towards children). Thus, we propose ClaimSpect, a retrieval-augmented generation-based framework for automatically constructing a hierarchy of aspects typically considered when addressing a claim and enriching them with corpus-specific perspectives. This structure hierarchically partitions an input corpus to retrieve relevant segments, which assist in discovering new sub-aspects. Moreover, these segments enable the discovery of varying perspectives towards an aspect of the claim (e.g., support, neutral, or oppose) and their respective prevalence (e.g., "how many biomedical papers believe vaccine A is more transportable than B?"). We apply ClaimSpect to a wide variety of real-world scientific and political claims featured in our constructed dataset, showcasing its robustness and accuracy in deconstructing a nuanced claim and representing perspectives within a corpus. Through real-world case studies and human evaluation, we validate its effectiveness over multiple baselines.
This paper presents our system for Track 1: Mistake Identification in the BEA 2025 Shared Task on Pedagogical Ability Assessment of AI-powered Tutors. The task involves evaluating whether a tutor's response correctly identifies a mistake in a student's mathematical reasoning. We explore four approaches: (1) an ensemble of machine learning models over pooled token embeddings from multiple pretrained language models (LMs); (2) a frozen sentence-transformer using [CLS] embeddings with an MLP classifier; (3) a history-aware model with multi-head attention between token-level history and response embeddings; and (4) a retrieval-augmented few-shot prompting system with a large language model (LLM) i.e. GPT 4o. Our final system retrieves semantically similar examples, constructs structured prompts, and uses schema-guided output parsing to produce interpretable predictions. It outperforms all baselines, demonstrating the effectiveness of combining example-driven prompting with LLM reasoning for pedagogical feedback assessment. Our code is available at https://github.com/NaumanNaeem/BEA_2025.
The advancement of conversational AI systems relies on the availability of high-quality, flexible, and reproducible synthetic dialogues for training, evaluation, and benchmarking. SDialog is a modular, extensible Python toolkit designed to address the challenges of synthetic dialogue generation and analysis. By leveraging instruction-tuned Large Language Models (LLMs), SDialog provides abstractions for personas, orchestration, and scenario management, enabling the creation of realistic, diverse, and controllable conversational data for research and development. SDialog supports workflows such as multi-agent simulation and scenario-driven generation, and represents a step forward in the standardization of tools and frameworks for synthetic data generation, a crucial advancement for ensuring reproducibility in today's fast-evolving research landscape.
We propose an unsupervised method for the reconstruction of protoforms i.e., ancestral word forms from which modern language forms are derived. While prior work has primarily relied on probabilistic models of phonological edits to infer protoforms from cognate sets, such approaches are limited by their predominantly data-driven nature. In contrast, our model integrates data-driven inference with rule-based heuristics within an evolutionary optimization framework. This hybrid approach leverages on both statistical patterns and linguistically motivated constraints to guide the reconstruction process. We evaluate our method on the task of reconstructing Latin protoforms using a dataset of cognates from five Romance languages. Experimental results demonstrate substantial improvements over established baselines across both character-level accuracy and phonological plausibility metrics.
Scientific claim verification against tables typically requires predicting whether a claim is supported or refuted given a table. However, we argue that predicting the final label alone is insufficient: it reveals little about the model's reasoning and offers limited interpretability. To address this, we reframe table-text alignment as an explanation task, requiring models to identify the table cells essential for claim verification. We build a new dataset by extending the SciTab benchmark with human-annotated cell-level rationales. Annotators verify the claim label and highlight the minimal set of cells needed to support their decision. After the annotation process, we utilize the collected information and propose a taxonomy for handling ambiguous cases. Our experiments show that (i) incorporating table alignment information improves claim verification performance, and (ii) most LLMs, while often predicting correct labels, fail to recover human-aligned rationales, suggesting that their predictions do not stem from faithful reasoning.
Recent advancements in Multimodal Emotion Recognition (MER) face challenges in addressing both modality missing and Out-Of-Distribution (OOD) data simultaneously. Existing methods often rely on specific models or introduce excessive parameters, which limits their practicality. To address these issues, we propose a novel robust MER framework, Causal Inference Distiller (CIDer), and introduce a new task, Random Modality Feature Missing (RMFM), to generalize the definition of modality missing. CIDer integrates two key components: a Model-Specific Self-Distillation (MSSD) module and a Model-Agnostic Causal Inference (MACI) module. MSSD enhances robustness under the RMFM task through a weight-sharing self-distillation approach applied across low-level features, attention maps, and high-level representations. Additionally, a Word-level Self-aligned Attention Module (WSAM) reduces computational complexity, while a Multimodal Composite Transformer (MCT) facilitates efficient multimodal fusion. To tackle OOD challenges, MACI employs a tailored causal graph to mitigate label and language biases using a Multimodal Causal Module (MCM) and fine-grained counterfactual texts. Notably, MACI can independently enhance OOD generalization with minimal additional parameters. Furthermore, we also introduce the new repartitioned MER OOD datasets. Experimental results demonstrate that CIDer achieves robust performance in both RMFM and OOD scenarios, with fewer parameters and faster training compared to state-of-the-art methods. The implementation of this work is publicly accessible at https://github.com/gw-zhong/CIDer.
Retrieval-Augmented Generation (RAG) has demonstrated considerable effectiveness in open-domain question answering. However, when applied to heterogeneous documents, comprising both textual and tabular components, existing RAG approaches exhibit critical limitations. The prevailing practice of flattening tables and chunking strategies disrupts the intrinsic tabular structure, leads to information loss, and undermines the reasoning capabilities of LLMs in multi-hop, global queries. To address these challenges, we propose TableRAG, an hybrid framework that unifies textual understanding and complex manipulations over tabular data. TableRAG iteratively operates in four steps: context-sensitive query decomposition, text retrieval, SQL programming and execution, and compositional intermediate answer generation. We also develop HeteQA, a novel benchmark designed to evaluate the multi-hop heterogeneous reasoning capabilities. Experimental results demonstrate that TableRAG consistently outperforms existing baselines on both public datasets and our HeteQA, establishing a new state-of-the-art for heterogeneous document question answering. We release TableRAG at https://github.com/yxh-y/TableRAG/tree/main.
Training large language models (LLMs) with chain-of-thought (CoT) supervision has proven effective for enhancing their reasoning abilities. However, obtaining reliable and accurate reasoning supervision remains a significant challenge. We propose a scalable method for generating a high-quality CoT supervision dataset by leveraging the determinism of program execution. Unlike existing reasoning dataset generation methods that rely on costly human annotations or error-prone LLM-generated CoT, our approach extracts verifiable, step-by-step reasoning traces from code execution and transforms them into a natural language CoT reasoning. Experiments on reasoning benchmarks across various domains show that our method effectively equips LLMs with transferable reasoning abilities across diverse tasks. Furthermore, the ablation studies validate that our method produces highly accurate reasoning data and reduces overall token length during inference by reducing meaningless repetition and overthinking.
This study examines how user-provided suggestions affect Large Language Models (LLMs) in a simulated educational context, where sycophancy poses significant risks. Testing five different LLMs from the OpenAI GPT-4o and GPT-4.1 model classes across five experimental conditions, we show that response quality varies dramatically based on query framing. In cases where the student mentions an incorrect answer, the LLM correctness can degrade by as much as 15 percentage points, while mentioning the correct answer boosts accuracy by the same margin. Our results also show that this bias is stronger in smaller models, with an effect of up to 30% for the GPT-4.1-nano model, versus 8% for the GPT-4o model. Our analysis of how often LLMs "flip" their answer, and an investigation into token level probabilities, confirm that the models are generally changing their answers to answer choices mentioned by students in line with the sycophancy hypothesis. This sycophantic behavior has important implications for educational equity, as LLMs may accelerate learning for knowledgeable students while the same tools may reinforce misunderstanding for less knowledgeable students. Our results highlight the need to better understand the mechanism, and ways to mitigate, such bias in the educational context.
Often when we interact with a first-person account of events, we consider whether or not the narrator, the primary speaker of the text, is reliable. In this paper, we propose using computational methods to identify unreliable narrators, i.e. those who unintentionally misrepresent information. Borrowing literary theory from narratology to define different types of unreliable narrators based on a variety of textual phenomena, we present TUNa, a human-annotated dataset of narratives from multiple domains, including blog posts, subreddit posts, hotel reviews, and works of literature. We define classification tasks for intra-narrational, inter-narrational, and inter-textual unreliabilities and analyze the performance of popular open-weight and proprietary LLMs for each. We propose learning from literature to perform unreliable narrator classification on real-world text data. To this end, we experiment with few-shot, fine-tuning, and curriculum learning settings. Our results show that this task is very challenging, and there is potential for using LLMs to identify unreliable narrators. We release our expert-annotated dataset and code and invite future research in this area.
Large language models (LLMs) excel at generating empathic responses in text-based conversations. But, how reliably do they judge the nuances of empathic communication? We investigate this question by comparing how experts, crowdworkers, and LLMs annotate empathic communication across four evaluative frameworks drawn from psychology, natural language processing, and communications applied to 200 real-world conversations where one speaker shares a personal problem and the other offers support. Drawing on 3,150 expert annotations, 2,844 crowd annotations, and 3,150 LLM annotations, we assess inter-rater reliability between these three annotator groups. We find that expert agreement is high but varies across the frameworks' sub-components depending on their clarity, complexity, and subjectivity. We show that expert agreement offers a more informative benchmark for contextualizing LLM performance than standard classification metrics. Across all four frameworks, LLMs consistently approach this expert level benchmark and exceed the reliability of crowdworkers. These results demonstrate how LLMs, when validated on specific tasks with appropriate benchmarks, can support transparency and oversight in emotionally sensitive applications including their use as conversational companions.
We investigate how large language models respond to prompts that differ only in their token-level realization but preserve the same semantic intent, a phenomenon we call prompt variance. We propose Prompt-Based Semantic Shift (PBSS), a diagnostic framework for measuring behavioral drift in LLMs under semantically equivalent prompt rewordings. Applied to ten constrained tasks, PBSS reveals consistent, model-specific response shifts, suggesting statistical regularities linked to tokenization and decoding. These results highlight an overlooked dimension of model evaluation stability under rephrasing and suggest that tokenization strategies and decoding dynamics may contribute to post-training quality of service instability.
Semantic degeneracy represents a fundamental property of natural language that extends beyond simple polysemy to encompass the combinatorial explosion of potential interpretations that emerges as semantic expressions increase in complexity. Large Language Models (LLMs) and other modern NLP systems face inherent limitations precisely because they operate within natural language itself, making them subject to the same interpretive constraints imposed by semantic degeneracy. In this work, we argue using Kolmogorov complexity that as an expression's complexity grows, the likelihood of any interpreting agent (human or LLM-powered AI) recovering the single intended meaning vanishes. This computational intractability suggests the classical view that linguistic forms possess meaning in and of themselves is flawed. We alternatively posit that meaning is instead actualized through an observer-dependent interpretive act. To test this, we conducted a semantic Bell inequality test using diverse LLM agents as ``computational cognitive systems'' to interpret ambiguous word pairs under varied contextual settings. Across several independent experiments, we found average CHSH expectation values ranging from 1.2 to 2.8, with several runs yielding values (e.g., 2.3-2.4) that significantly violate the classical boundary ($|S|\leq2$). This demonstrates that linguistic interpretation under ambiguity can exhibit non-classical contextuality, consistent with results from human cognition experiments. These results inherently imply that classical frequentist-based analytical approaches for natural language are necessarily lossy. Instead, we propose that Bayesian-style repeated sampling approaches can provide more practically useful and appropriate characterizations of linguistic meaning in context.
Agentic tasks, which require multi-step problem solving with autonomy, tool use, and adaptive reasoning, are becoming increasingly central to the advancement of NLP and AI. However, existing instruction data lacks tool interaction, and current agentic benchmarks rely on costly human annotation, limiting their scalability. We introduce \textsc{TaskCraft}, an automated workflow for generating difficulty-scalable, multi-tool, and verifiable agentic tasks with execution trajectories. TaskCraft expands atomic tasks using depth-based and width-based extensions to create structurally and hierarchically complex challenges. Empirical results show that these tasks improve prompt optimization in the generation workflow and enhance supervised fine-tuning of agentic foundation models. We present a large-scale synthetic dataset of approximately 36,000 tasks with varying difficulty to support future research on agent tuning and evaluation.
A modern digital pathology vendor-agnostic binary slide format specifically targeting the unmet need of efficient real-time transfer and display has not yet been established. The growing adoption of digital pathology only intensifies the need for an intermediary digital slide format that emphasizes performance for use between slide servers and image management software. The DICOM standard is a well-established format widely used for the long-term storage of both images and associated critical metadata. However, it was inherently designed for radiology rather than digital pathology, a discipline that imposes a unique set of performance requirements due to high-speed multi-pyramidal rendering within whole slide viewer applications. Here we introduce the Iris file extension, a binary container specification explicitly designed for performance-oriented whole slide image viewer systems. The Iris file extension specification is explicit and straightforward, adding modern compression support, a dynamic structure with fully optional metadata features, computationally trivial deep file validation, corruption recovery capabilities, and slide annotations. In addition to the file specification document, we provide source code to allow for (de)serialization and validation of a binary stream against the standard. We also provide corresponding binary builds with C++, Python, and JavaScript language support. Finally, we provide full encoder and decoder implementation source code, as well as binary builds (part of the separate Iris Codec Community module), with language bindings for C++ and Python, allowing for easy integration with existing WSI solutions. We provide the Iris File Extension specification openly to the community in the form of a Creative Commons Attribution-No Derivative 4.0 International license.
Vision-Language Models (VLMs) have shown remarkable performance on diverse visual and linguistic tasks, yet they remain fundamentally limited in their understanding of 3D spatial structures. We propose Geometric Distillation, a lightweight, annotation-free fine-tuning framework that injects human-inspired geometric cues into pretrained VLMs without modifying their architecture. By distilling (1) sparse correspondences, (2) relative depth relations, and (3) dense cost volumes from off-the-shelf 3D foundation models (e.g., MASt3R, VGGT), our method shapes representations to be geometry-aware while remaining compatible with natural image-text inputs. Through extensive evaluations on 3D vision-language reasoning and 3D perception benchmarks, our method consistently outperforms prior approaches, achieving improved 3D spatial reasoning with significantly lower computational cost. Our work demonstrates a scalable and efficient path to bridge 2D-trained VLMs with 3D understanding, opening up wider use in spatially grounded multimodal tasks.
Large Reasoning Models (LRMs) like o1 and DeepSeek-R1 have shown remarkable progress in natural language reasoning with long chain-of-thought (CoT), yet they remain inefficient or inaccurate when handling complex mathematical operations. Addressing these limitations through computational tools (e.g., computation libraries and symbolic solvers) is promising, but it introduces a technical challenge: Code Interpreter (CI) brings external knowledge beyond the model's internal text representations, thus the direct combination is not efficient. This paper introduces CoRT, a post-training framework for teaching LRMs to leverage CI effectively and efficiently. As a first step, we address the data scarcity issue by synthesizing code-integrated reasoning data through Hint-Engineering, which strategically inserts different hints at appropriate positions to optimize LRM-CI interaction. We manually create 30 high-quality samples, upon which we post-train models ranging from 1.5B to 32B parameters, with supervised fine-tuning, rejection fine-tuning and reinforcement learning. Our experimental results demonstrate that Hint-Engineering models achieve 4\% and 8\% absolute improvements on DeepSeek-R1-Distill-Qwen-32B and DeepSeek-R1-Distill-Qwen-1.5B respectively, across five challenging mathematical reasoning datasets. Furthermore, Hint-Engineering models use about 30\% fewer tokens for the 32B model and 50\% fewer tokens for the 1.5B model compared with the natural language models. The models and code are available at https://github.com/ChengpengLi1003/CoRT.
Intermediate Respiratory Care Units (IRCUs) are vital during crises like COVID-19. This study evaluated clinical outcomes and operational dynamics of a new Spanish IRCU with specialized staffing. A prospective cohort study (April-August 2021) included 249 adult patients with COVID-19 respiratory failure (UHVN IRCU, Granada). Data on demographics, Non-Invasive Ventilation (NIV), length of stay (LOS), and outcomes (ICU transfer, exitus, recovery) were analyzed. Patient flow was simulated using a data-calibrated deterministic compartmental model (Ordinary Differential Equations, ODEs) that represented state transitions, and an empirical LOS-based stochastic convolution model that incorporated admission variability. The median age was 51; 31% of patients required NIV. NIV patients were older (median 61 vs 42, p<0.001). Overall, 8% needed ICU transfer; 3% experienced in-IRCU exitus. Notably, no ICU transfers or deaths occurred among 172 non-NIV patients. Of 77 high-risk NIV patients, 68% recovered in IRCU without ICU escalation. The ODE model, based on transition rates between patient states, reflected aggregate outcomes. Both modeling approaches demonstrated system strain during admission surges (partially mitigated by simulated care efficiency improvements via parameter modulation) and yielded consistent peak occupancy estimates. This IRCU, with specialized staffing, effectively managed severe COVID-19. High recovery rates, especially for NIV patients, potentially eased ICU pressure. Dynamic modeling confirmed surge vulnerability but highlighted the benefits of care efficiency from modulated transition parameters. Findings underscore positive outcomes in this IRCU model and support such units in pandemic response.
The advent of Large Weather Models (LWMs) has marked a turning point in data-driven forecasting, with many models now outperforming traditional numerical systems in the medium range. However, achieving stable, long-range autoregressive forecasts beyond a few weeks remains a significant challenge. Prevailing state-of-the-art models that achieve year-long stability, such as SFNO and DLWP-HPX, have relied on transforming input data onto non-standard spatial domains like spherical harmonics or HEALPix meshes. This has led to the prevailing assumption that such representations are necessary to enforce physical consistency and long-term stability. This paper challenges that assumption by investigating whether comparable long-range performance can be achieved on the standard latitude-longitude grid. We introduce AtmosMJ, a deep convolutional network that operates directly on ERA5 data without any spherical remapping. The model's stability is enabled by a novel Gated Residual Fusion (GRF) mechanism, which adaptively moderates feature updates to prevent error accumulation over long recursive simulations. Our results demonstrate that AtmosMJ produces stable and physically plausible forecasts for about 500 days. In quantitative evaluations, it achieves competitive 10-day forecast accuracy against models like Pangu-Weather and GraphCast, all while requiring a remarkably low training budget of 5.7 days on a V100 GPU. Our findings suggest that efficient architectural design, rather than non-standard data representation, can be the key to unlocking stable and computationally efficient long-range weather prediction.
This paper presents a system developed for SemEval 2025 Task 8: Question Answering (QA) over tabular data. Our approach integrates several key components: text-to-SQL and text-to-code generation modules, a self-correction mechanism, and a retrieval-augmented generation (RAG). Additionally, it includes an end-to-end (E2E) module, all orchestrated by a large language model (LLM). Through ablation studies, we analyzed the effects of different parts of our pipeline and identified the challenges that are still present in this field. During the evaluation phase of the competition, our solution achieved an accuracy of 80%, resulting in a top-13 ranking among the 38 participating teams. Our pipeline demonstrates a significant improvement in accuracy for open-source models and achieves a performance comparable to proprietary LLMs in QA tasks over tables. The code is available at GitHub repository.