Successful multi-turn voice interaction requires models to retain information from earlier turns, follow persistent instructions, and track revisions to conversational state. These abilities are usually evaluated from transcripts, leaving open whether successes demonstrated from text carry over when the same turns are spoken, and whether this transfer is consistent across languages. We introduce Omnilingua-AudioMC, a strictly parallel multilingual extension of Audio MultiChallenge, to study this question. The benchmark contains 354 conversations and 1,404 rubric items across ten languages, with content, turn structure, cross-turn dependencies, and evaluation criteria aligned across transcript and speech conditions. Across six speech-language models, the shift from reference transcripts to direct speech input is strongly model- and language-dependent: models with similar transcript performance can diverge substantially under speech, and their relative strengths across languages can change. Paired rubric-level analysis shows that aggregate transcript-speech gaps can obscure which conversational requirements are preserved, because losses of transcript-demonstrated successes can be offset by gains on other items. Replacing preceding user turns, the current turn, or both with model-generated transcripts localizes recoverable degradation: for the two models with the largest gaps, history substitution yields a larger macro-level improvement than current-turn substitution in all ten languages, while the strongest direct-speech model gains little from either. These results show that transcript performance is an unreliable proxy for multilingual spoken conversational ability and identify the representation of preceding spoken context as an important source of recoverable degradation. Our code is publicly available at https://github.com/QwenLM/Omnilingua-Bench/tree/main/Omnilingua-AudioMC.
Omni-modal large language models (LLMs) have achieved remarkable progress in audio and video understanding. However, existing models struggle with long-form, multi-speaker videos, especially in understanding the entire video and recognizing who says what and when. To characterize the capability limit of current omni LLMs and guide future model development, we introduce Omnilingual-MsASR, a multilingual, multimodal benchmark for long-form Multi-speaker Automatic Speech Recognition (MsASR) covering six languages, recordings up to 30 minutes, and scenarios with up to 20 speakers. Omnilingual-MsASR provides parallel audio and video inputs to evaluate models' cross-modal consistency and their ability to exploit complementary cues across modalities. Extensive evaluations of leading omni-modal LLMs reveal substantial performance bottlenecks, while their tcpWERs frequently exceed 100%. These findings highlight fine-grained temporal localization, speaker disambiguation, and cross-modal synergy as key challenges for future omni-modal foundation models. Our work is available at https://github.com/QwenLM/Omnilingua-Bench/tree/main/Omnilingua-MSpeaker.
Zipei Zhang, Jaakko Sahlsten, Mikko Saukkoriipi, Johanna Ronkainen, Mona Lehto, Akseli Aittoniemi, Irina Rinta-Kiikka, Kimmo Kaski
Assessment of the main pancreatic duct (MPD) from volumetric magnetic resonance cholangiopancreatography (MRCP) scans is an essential task for ductal dilatation diagnosis and pancreatic cancer analysis. However, manual assessment is affected by reader variability in which automated analysis may assist. Nevertheless, fully automated and uncertainty-aware MPD analysis from volumetric MRCP remains underexplored area of research. In this study, we propose an automatic and uncertainty-aware deep learning-based pipeline for MPD segmentation and maximum-width analysis from MRCP scans. A coarse localisation model first identifies the ductal region-of-interest, followed by a fine model that segments the MPD within the region. The dataset consists of 256 MPD-segmented MRCP scans from 69 patients. A randomly selected holdout test subset of 14 patients with 55 MRCP volumes was used for evaluation. We employed a five-model cross-validation ensemble for improved performance and uncertainty estimation. The best MPD segmentation performance was 0.634 DSC, 0.744 precision, and 0.593 recall. The mean absolute error for maximum MPD width prediction was 0.355~mm and dilated MPD ($\geq 5$~mm) was detected with 75.0\% sensitivity and 74.4\% specificity. These novel results suggest that automated MPD segmentation can be used to aid MRCP-based duct measurements.
Kaibo Huang, Jin Tan, Yukun Wei, Wanling Li, Zipei Zhang, Hui Tian Huaqiao University, Zhongliang Yang Beijing University of Posts and Telecommunications, Linna Zhou Beijing University of Posts and Telecommunications
LLM-based agents are increasingly deployed to autonomously solve complex tasks, raising urgent needs for IP protection and regulatory provenance. While content watermarking effectively attributes LLM-generated outputs, it fails to directly identify the high-level planning behaviors (e.g., tool and subgoal choices) that govern multi-step execution. Critically, watermarking at the planning-behavior layer faces unique challenges: minor distributional deviations in decision-making can compound during long-term agent operation, degrading utility, and many agents operate as black boxes that are difficult to intervene in directly. To bridge this gap, we propose AgentMark, a behavioral watermarking framework that embeds multi-bit identifiers into planning decisions while preserving utility. It operates by eliciting an explicit behavior distribution from the agent and applying distribution-preserving conditional sampling, enabling deployment under black-box APIs while remaining compatible with action-layer content watermarking. Experiments across embodied, tool-use, and social environments demonstrate practical multi-bit capacity, robust recovery from partial logs, and utility preservation. Code is available at https://github.com/Tooooa/AgentMark.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao 等 60 位作者
In this work, we present Qwen3, the latest version of the Qwen model family. Qwen3 comprises a series of large language models (LLMs) designed to advance performance, efficiency, and multilingual capabilities. The Qwen3 series includes models of both dense and Mixture-of-Expert (MoE) architectures, with parameter scales ranging from 0.6 to 235 billion. A key innovation in Qwen3 is the integration of thinking mode (for complex, multi-step reasoning) and non-thinking mode (for rapid, context-driven responses) into a unified framework. This eliminates the need to switch between different models--such as chat-optimized models (e.g., GPT-4o) and dedicated reasoning models (e.g., QwQ-32B)--and enables dynamic mode switching based on user queries or chat templates. Meanwhile, Qwen3 introduces a thinking budget mechanism, allowing users to allocate computational resources adaptively during inference, thereby balancing latency and performance based on task complexity. Moreover, by leveraging the knowledge from the flagship models, we significantly reduce the computational resources required to build smaller-scale models, while ensuring their highly competitive performance. Empirical evaluations demonstrate that Qwen3 achieves state-of-the-art results across diverse benchmarks, including tasks in code generation, mathematical reasoning, agent tasks, etc., competitive against larger MoE models and proprietary models. Compared to its predecessor Qwen2.5, Qwen3 expands multilingual support from 29 to 119 languages and dialects, enhancing global accessibility through improved cross-lingual understanding and generation capabilities. To facilitate reproducibility and community-driven research and development, all Qwen3 models are publicly accessible under Apache 2.0.