Nikita Afonin, Nikita Andriianov, Vahagn Hovhannisyan, Nikhil Bageshpura, Kyle Liu, Kevin Zhu, Sunishchal Dev, Ashwinee Panda, Oleg Rogov, Elena Tutubalina, Alexander Panchenko, Mikhail Seleznyov. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Current large language models (LLMs) struggle to answer questions that span tens of thousands of tokens, especially when multi-hop reasoning is involved.While prior benchmarks explore long-context comprehension or multihop reasoning in isolation, none jointly vary context length and reasoning depth in natural narrative settings.We introduce NOVEL-HOPQA, the first benchmark to evaluate 1-4 hop QA over 64k-128k-token excerpts from 83 full-length public-domain novels.A keywordguided pipeline builds hop-separated chains grounded in coherent storylines.We evaluate seven state-of-the-art (SOTA) models and apply oracle-context filtering to ensure all questions are genuinely answerable.Human annotators validate both alignment and hop depth.We additionally present retrieval-augmented generation (RAG) evaluations to test model performance when only selected passages are provided instead of the full context.We noticed consistent accuracy drops with increased hops and context length, even in frontier models-revealing that sheer scale does not guarantee robust reasoning.Our failure mode analysis highlights common breakdowns, such as missed final-hop integration and long-range drift.NOVELHOPQA offers a controlled diagnostic setting to test multi-hop reasoning at scale.All code and datasets are available at: https://novelhopqa.github.io.

We introduce MusclePose as an end-to-end learnable physics-infused 3D human pose estimator that incorporates muscle-dynamics modeling to infer human dynamics from monocular video. Current physics pose estimators aim to predict physically plausible poses by enforcing the underlying dynamics equations that govern motion. Since this is an underconstrained problem without force-annotated data, methods often estimate kinetics with external physics optimizers that may not be compatible with existing learning frameworks, or are too slow for real-time inference. While more recent methods use a regression-based approach to overcome these issues, the estimated kinetics can be seen as auxiliary predictions, and may not be physically plausible. To this end, we build on existing regression-based approaches, and aim to improve the biofidelity of kinetic inference with a multihypothesis approach --- by inferring joint torques via Lagrange’s equations and via muscle dynamics modeling with muscle torque generators. Furthermore, MusclePose predicts detailed human anthropometrics based on values from biomechanics studies, in contrast to existing physics pose estimators that construct their human models with shape primitives. We show that MusclePose is competitive with existing 3D pose estimators in positional accuracy, while also able to infer plausible human kinetics and muscle signals consistent with values from biomechanics studies, without requiring an external physics engine.

Grounded in the diffusion of innovation theory and the technology–organization–environment framework, we develop an integrative model to study the determinants of post-adoption stages of innovation diffusion, using enterprise digital transformation as an example of technology-enabled innovations. We specify four innovation characteristics (relative advantage, compatibility, costs and security concern) and four contextual factors (technology competence, organization size, competitive pressure and partner readiness) as determinants of post-adoption usage, and postulate usage as an intermediate link to impact on firm performance. We test the proposed model using a dataset of 1415 companies from six European countries. We find that the innovation needs to be used extensively in value-chain activities before its impact can be realized. Among the innovation characteristics, we find that compatibility is the strongest driver, and security concern outweighs cost as a usage inhibitor. Among the contextual variables, technology competence, partner readiness and competitive pressure significantly drive e-business usage, and the structural inertia of large firms tends to slow down its penetration. Collectively, these results indicate that innovation diffusion can be better understood by including both innovation characteristics and contextual factors, whereas earlier literature has traditionally treated the two separately. Finally, we evaluate an international dimension among European countries and tease out important boundary conditions that would not have been evident in a single-country dataset. Our results show that careful attention must be paid to the economic and regulatory factors that may result in uneven innovation diffusion even among developed European countries.

In this study, we developed a conceptual model for studying the adoption of electronic business (e-business or EB) at the firm level, incorporating six adoption facilitators and inhibitors, based on the technology–organization–environment theoretical framework. Survey data from 3100 businesses and 7500 consumers in eight European countries were used to test the proposed adoption model. We conducted confirmatory factor analysis to assess the reliability and validity of constructs. To examine whether adoption patterns differ across different e-business environments, we divided the full sample into high EB-intensity and low EB-intensity countries. After controlling for variations of industry and country effects, the fitted logit models demonstrated four findings: (1) Technology competence, firm scope and size, consumer readiness, and competitive pressure are significant adoption drivers, while lack of trading partner readiness is a significant adoption inhibitor. (2) As EB-intensity increases, two environmental factors – consumer readiness and lack of trading partner readiness – become less important, while competitive pressure remains significant. (3) In high EB-intensity countries, e-business is no longer a phenomenon dominated by large firms; as more and more firms engage in e-business, network effect works to the advantage of small firms. (4) Firms are more cautious in adopting e-business in high EB-intensity countries – it seems to suggest that the more informed firms are less aggressive in adopting e-business, a somehow surprising result. Explanations and implications are offered.