While Large Language Models have significantly advanced Text2SQL generation, a critical semantic gap persists: syntactically valid queries can still misinterpret user intent. To mitigate this challenge, we propose GBV-SQL, a multi-agent framework that introduces Guided Generation with SQL2Text Back-translation Validation. In particular, a dedicated validator translates generated SQL back into natural language and checks whether its logic is aligned with the original question. Beyond the method itself, we also conduct a systematic audit of benchmark quality and introduce a typology of “Gold Errors” in Text2SQL datasets. Our analysis shows that benchmark issues can coexist with strong execution accuracy and can substantially affect evaluation outcomes. On the challenging BIRD benchmark, GBV-SQL achieves 63.23% execution accuracy, a 5.8% absolute improvement over the Deepseek-v3-based MAC-SQL setting. Under manually audited benchmark corrections, GBV-SQL reaches 96.5% (dev) and 97.6% (test) on Spider, and 90.42% on repaired BIRD dev, providing a diagnostic view of model behavior under improved gold quality. Our work contributes both a practical framework for semantic validation and an empirical analysis of benchmark integrity in Text2SQL evaluation.

Large language models (LLMs) have achieved strong performance on standard benchmarks, yet their performance is not robust across different task manifestations. It remains unclear how performance changes under controlled task rewrites that preserve the original solution structure, while varying the rewrite type and level. To address this question, we introduce ReTRE (Rewrite-based Transfer Robustness Evaluation), an evaluation benchmark inspired by learning transfer theory that probes transfer robustness along two rewrite levels: Near Transfer and Far Transfer. ReTRE employs a multi-agent system to construct textual and visual variants while preserving the structure of the original solution. Evaluations on mathematical and science tasks across state-of-the-art multimodal LLMs reveal a consistent transfer gap: performance exhibits a general declining trend as transfer similarity drops and strong text performance can face performance decline under cross-modal transfer. Crucially, we identify a divergence between post-training paradigms: reinforcement learning preserves transfer robustness, whereas supervised fine-tuning tends to overfit the training distribution, leading to severe degradation in far-transfer performance despite strong in-distribution accuracy.

Question answering over temporal knowledge graphs (TKGQA) is an emerging topic, which has attracted increasing interest since it considers the dynamic knowledge in the world. Several datasets along with model developments are proposed in the TKGQA research field. However, existing studies generally focus on fact-centered reasoning, with limited attention to temporal reasoning. To tackle the intricate and comprehensive nature of temporal reasoning, we propose a new TKGQA dataset, MusTQ, which contains 666K multi-step temporal reasoning questions as well as a TKG. The multi-step temporal reasoning is established based on six basic temporal reasoning types derived from a well-established measure theory. Using MusTQ, we evaluate previous TKGQA methods and find that they typically fall short in multi-step temporal reasoning. Furthermore, we propose a TKGQA model, MusTKGQA, which enhances multi-step reasoning ability with entity-time attention mechanism and optimized temporal knowledge graph representation. Extensive experiments on MusTQ show that our model achieves state-of-the-art multi-step temporal reasoning performance.

Despite advancements in Large Language Models (LLMs) and Large Multimodal Models (LMMs), their integration into language-grounded, human-like embodied agents remains incomplete, hindering complex real-life task performance in 3D environments. Existing integrations often feature limited open-sourcing, challenging collective progress in this field. We introduce LEGENT, an open, scalable platform for developing embodied agents using LLMs and LMMs. LEGENT offers a dual approach: a rich 3D environment with interactive, communicable, and actionable agents, paired with a user-friendly interface, and a sophisticated data generation pipeline utilizing advanced algorithms to exploit supervision from simulated worlds at scale. In our experiments, an embryonic vision-language-action model trained on LEGENT-generated data surpasses GPT-4V in embodied tasks, showcasing promising generalization capabilities. The demo video is available at the following link https://video.legent.ai.

While Large language models (LLMs) have demonstrated considerable capabilities across various natural language tasks, they often fall short of the performance achieved by domain-specific state-of-the-art models. One potential approach to enhance domain-specific capabilities of LLMs involves fine-tuning them using corresponding datasets. However, this method can be both resource and time-intensive, and not applicable to closed-source commercial LLMs. In this paper, we propose Preference Adaptation for Enhancing Domain-specific Abilities of LLMs (PANDA), a method designed to augment the domain-specific capabilities of LLMs by leveraging insights from the response preference of expert models without requiring fine-tuning. Our experimental results reveal that PANDA significantly enhances the domain-specific ability of LLMs on text classification and interactive decision tasks. Moreover, LLM with PANDA even outperforms the expert model that being learned on 4 tasks of ScienceWorld. This finding highlights the potential of exploring tuning-free approaches to achieve weak-to-strong generalization.

Zero pronoun resolution aims at recognizing dropped pronouns and pointing out their anaphoric mentions, while non-zero coreference resolution targets at clustering mentions referring to the same entity. Existing efforts often deal with the two problems separately regardless of their close essential correlations. In this paper, we investigate the possibility of jointly solving zero pronoun resolution and coreference resolution via a novel end-to-end neural model. Specifically, we design a gap-masked self-attention model that encodes gaps and tokens in the same space, where gaps could capture valuable contextual information according to their surrounding tokens while tokens could maintain original sequential information without disturbance. Additionally, we also propose a two-stage interaction mechanism to make full use of the exclusive relationship between zero pronouns and mentions. Our empirical study conducted on the OntoNotes 5.0 Chinese dataset shows that our model could outperform corresponding state-of-the-art approaches on both tasks.