Preference-based reinforcement learning (PbRL) enables agents to learn control policies without requiring manually designed reward functions, making it well-suited for tasks where objectives are difficult to formalize or inherently subjective. Acrobatic flight poses a particularly challenging problem due to its complex dynamics, rapid movements, and the importance of precise execution. In this work, we explore the use of PbRL for agile drone control, focusing on the execution of dynamic maneuvers such as powerloops. Building on Preference-based Proximal Policy Optimization (Preference PPO), we propose Reward Ensemble under Confidence (REC), an extension to the reward learning objective that improves preference modeling and learning stability. Our method achieves 88.4% of the shaped reward performance, compared to 55.2% with standard Preference PPO. We train policies in simulation and successfully transfer them to real-world drones, demonstrating multiple acrobatic maneuvers where human preferences emphasize stylistic qualities of motion. Furthermore, we demonstrate the applicability of our probabilistic reward model in a representative MuJoCo environment for continuous control. Finally, we highlight the limitations of manually designed rewards, observing only 60.7% agreement with human preferences. These results underscore the effectiveness of PbRL in capturing complex, human-centered objectives across both physical and simulated domains.

While modern recommender systems are instrumental in navigating information abundance, they remain fundamentally limited by static user modeling and reactive decision-making paradigms. Current large language model (LLM)-based agents inherit these shortcomings through their overreliance on heuristic pattern matching, yielding recommendations prone to shallow correlation bias, limited causal inference, and brittleness in sparse-data scenarios. We introduce STARec, a slow-thinking augmented agent framework that endows recommender systems with autonomous deliberative reasoning capabilities. Each user is modeled as an agent with parallel cognitions: fast response for immediate interactions and slow reasoning that performs chain-of-thought rationales. To cultivate intrinsic slow thinking, we develop anchored reinforcement training - a two-stage paradigm combining structured knowledge distillation from advanced reasoning models with preference-aligned reward shaping. This hybrid approach scaffolds agents in acquiring foundational capabilities (preference summarization, rationale generation) while enabling dynamic policy adaptation through simulated feedback loops. Experiments on MovieLens 1M and Amazon CDs benchmarks demonstrate that STARec achieves substantial performance gains compared with state-of-the-art baselines, despite using only 0.4% of the full training data.

Grammatical error correction is a significant task in NLP. Traditional methods based on encoder-decoder models have achieved certain success, but the application of LLMs in this field is still underexplored. Current research predominantly relies on supervised fine-tuning to train LLMs to directly generate the corrected sentence, which limits the model's powerful reasoning ability. To address this limitation, we propose a novel framework based on Rule-Based RL. Through experiments on the Chinese datasets, our Rule-Based RL framework achieves \textbf{state-of-the-art }performance, with a notable increase in \textbf{recall}. This result clearly highlights the advantages of using RL to steer LLMs, offering a more controllable and reliable paradigm for future development in GEC.

Large language models (LLMs) with chain-of-thought reasoning have demonstrated remarkable problem-solving capabilities, but controlling their computational effort remains a significant challenge for practical deployment. Recent proprietary systems like OpenAI's gpt-oss series have introduced discrete operational modes for intuitive reasoning control, but the open-source community has largely failed to achieve such capabilities. In this paper, we introduce ThinkDial, the first open-recipe end-to-end framework that successfully implements gpt-oss-style controllable reasoning through discrete operational modes. Our system enables seamless switching between three distinct reasoning regimes: High mode (full reasoning capability), Medium mode (50 percent token reduction with <10 percent performance degradation), and Low mode (75 percent token reduction with <15 percent performance degradation). We achieve this through an end-to-end training paradigm that integrates budget-mode control throughout the entire pipeline: budget-mode supervised fine-tuning that embeds controllable reasoning capabilities directly into the learning process, and two-phase budget-aware reinforcement learning with adaptive reward shaping. Extensive experiments demonstrate that ThinkDial achieves target compression-performance trade-offs with clear response length reductions while maintaining performance thresholds. The framework also exhibits strong generalization capabilities on out-of-distribution tasks.

Multiple-choice questions (MCQs) play a crucial role in fostering deep thinking and knowledge integration in education. However, previous research has primarily focused on generating MCQs with textual options, but it largely overlooks the visual options. Moreover, generating high-quality distractors remains a major challenge due to the high cost and limited scalability of manual authoring. To tackle these problems, we propose a Cross-modal Options Synthesis (CmOS), a novel framework for generating educational MCQs with visual options. Our framework integrates Multimodal Chain-of-Thought (MCoT) reasoning process and Retrieval-Augmented Generation (RAG) to produce semantically plausible and visually similar answer and distractors. It also includes a discrimination module to identify content suitable for visual options. Experimental results on test tasks demonstrate the superiority of CmOS in content discrimination, question generation and visual option generation over existing methods across various subjects and educational levels.

This paper investigates the enhancement of reasoning capabilities in language models through token-level multi-model collaboration. Our approach selects the optimal tokens from the next token distributions provided by multiple models to perform autoregressive reasoning. Contrary to the assumption that more models yield better results, we introduce a distribution distance-based dynamic selection strategy (DDS) to optimize the multi-model collaboration process. To address the critical challenge of vocabulary misalignment in multi-model collaboration, we propose the concept of minimal complete semantic units (MCSU), which is simple yet enables multiple language models to achieve natural alignment within the linguistic space. Experimental results across various benchmarks demonstrate the superiority of our method. The code will be available at https://github.com/Fanye12/DDS.

Efficient querying and analysis of large tabular datasets remain significant challenges, especially for users without expertise in programming languages like SQL. Text-to-SQL approaches have shown promising performance on benchmark data; however, they inherit SQL's drawbacks, including inefficiency with large datasets and limited support for complex data analyses beyond basic querying. We propose a novel framework that transforms natural language queries into query plans. Our solution is implemented outside traditional databases, allowing us to support classical SQL commands while avoiding SQL's inherent limitations. Additionally, we enable complex analytical functions, such as principal component analysis and anomaly detection, providing greater flexibility and extensibility than traditional SQL capabilities. We leverage LLMs to iteratively interpret queries and construct operation sequences, addressing computational complexity by incrementally building solutions. By executing operations directly on the data, we overcome context length limitations without requiring the entire dataset to be processed by the model. We validate our framework through experiments on both standard databases and large scientific tables, demonstrating its effectiveness in handling extensive datasets and performing sophisticated data analyses.

Recent advances in prompt optimization, exemplified by methods such as TextGrad, enable automatic, gradient-like refinement of textual prompts to enhance the performance of large language models (LLMs) on specific downstream tasks. However, current approaches are typically stateless and operate independently across optimization runs, lacking mechanisms to preserve and leverage historical optimization experience. Furthermore, they are susceptible to overfitting, often yielding prompt updates that generalize poorly beyond the immediate task context. To address these limitations, we propose Reflection-Enhanced Meta-Optimization (REMO), a novel framework that integrates (1) a memory-augmented Reflection Retrieval-Augmented Generation (RAG) module - structured as a "mistake notebook" and (2) a Self-Adaptive Optimizer, implemented via an LLM-driven meta-controller that synthesizes epoch-level reflective insights to iteratively improve system-level prompting strategies. This architecture enables not only local, fine-grained prompt tuning akin to TextGrad, but also the systematic accumulation and reuse of cross-run optimization knowledge, thereby supporting continual improvement over time. We instantiate the REMO framework using Qwen3-32B in standard inference mode - without explicit chain-of-thought prompting - and evaluate its efficacy on the GSM8K benchmark for mathematical reasoning. Experimental results demonstrate that, compared to a TextGrad baseline, REMO achieves more stable and robust generalization, albeit at the cost of increased computational overhead. We provide a detailed exposition of the algorithmic design, conduct a qualitative and quantitative analysis of optimization dynamics, and present a comprehensive ablation study to elucidate the contributions of each component.

Emotion Cause Triplet Extraction in Multimodal Conversations (MECTEC) has recently gained significant attention in social media analysis, aiming to extract emotion utterances, cause utterances, and emotion categories simultaneously. However, the scarcity of related datasets, with only one published dataset featuring highly uniform dialogue scenarios, hinders model development in this field. To address this, we introduce MECAD, the first multimodal, multi-scenario MECTEC dataset, comprising 989 conversations from 56 TV series spanning a wide range of dialogue contexts. In addition, existing MECTEC methods fail to explicitly model emotional and causal contexts and neglect the fusion of semantic information at different levels, leading to performance degradation. In this paper, we propose M3HG, a novel model that explicitly captures emotional and causal contexts and effectively fuses contextual information at both inter- and intra-utterance levels via a multimodal heterogeneous graph. Extensive experiments demonstrate the effectiveness of M3HG compared with existing state-of-the-art methods. The codes and dataset are available at https://github.com/redifinition/M3HG.

The generation of ad headlines plays a vital role in modern advertising, where both quality and diversity are essential to engage a broad range of audience segments. Current approaches primarily optimize language models for headline quality or click-through rates (CTR), often overlooking the need for diversity and resulting in homogeneous outputs. To address this limitation, we propose DIVER, a novel framework based on large language models (LLMs) that are jointly optimized for both diversity and quality. We first design a semantic- and stylistic-aware data generation pipeline that automatically produces high-quality training pairs with ad content and multiple diverse headlines. To achieve the goal of generating high-quality and diversified ad headlines within a single forward pass, we propose a multi-stage multi-objective optimization framework with supervised fine-tuning (SFT) and reinforcement learning (RL). Experiments on real-world industrial datasets demonstrate that DIVER effectively balances quality and diversity. Deployed on a large-scale content-sharing platform serving hundreds of millions of users, our framework improves advertiser value (ADVV) and CTR by 4.0% and 1.4%.