Bowen Zhou

Scientific intelligence, which requires long-horizon reasoning and research across diverse and highly specialized domains, remains a key challenge for artificial intelligence. Although large language models (LLMs) have achieved strong performance on general tasks, their single-model paradigm faces fundamental limitations in scientific domains, including reduced effectiveness under specialization, high scaling costs with diminishing returns, and insufficient capacity for fine-grained long-horizon tasks. In this paper, we introduce an extended-mind-inspired paradigm, extending intelligence to an agentic system composed of LLM, interactive objects, and autonomous interaction processes, enabling deep specialization, new scaling dimensions, and stronger representational capacity. Building on this, we introduce ExoMind, the first extended-mind-inspired agentic system, which integrate systematic data engineering, scientific interaction framework, and systematic training strategy. Across diverse scientific reasoning and research tasks, our method achieves substantial and consistent performance improvements using less data, small model, and low-cost training. Notably, ExoMind achieves leading performance among frontier open-source and closed-source models on challenging scientific benchmarks (as shown in Figure 1), while also improving general-purpose capabilities. These results suggest a customizable, cost-efficient, and high-performance pathway toward democratizing scientific intelligence.

Speculative decoding accelerates autoregressive large language model inference by producing low-cost draft work and committing accepted portions after parallel target-side verification. This draft-then-verify process defines the broad family, while recent methods differ in where draft work comes from, how candidate units are logically organized across drafting and verification, what fidelity claim licenses commitment, and how runtime systems realize committed state, control speculation, execute serving paths, and evaluate end-to-end performance. We therefore develop a systems co-design taxonomy with four interdependent aspects---Draft Source, Draft Geometry, Verification Fidelity, and Runtime Execution---and 19 categories. Applying it to 304 works yields Method Profiles that trace individual works through the design space and help researchers navigate the literature, compare methods consistently, and select category-matched baselines. We synthesize recurring combinations and interactions among the four aspects into Co-Design Gates---Representability, Verifiability, State Realizability, and Profitability---that organize open challenges and future directions. Together, the taxonomy, Method Profiles, and Gates frame speculative decoding as a systems co-design problem in which these choices must align to deliver the stated fidelity and end-to-end benefit.

Large language models (LLMs) have taken a great step towards AGI. Meanwhile, an increasing number of domain-specific problems such as math and programming boost these general-purpose models to continuously evolve via learning deeper expertise. Now is thus the time further to extend the diversity of specialized applications for knowledgeable LLMs, though collecting high quality data with unexpected and informative tasks is challenging. In this paper, we propose to use advertisement (ad) videos as a challenging test-bed to probe the ability of LLMs in perceiving beyond the objective physical content of common visual domain. Our motivation is to take full advantage of the clue-rich and information-dense ad videos' traits, e.g., marketing logic, persuasive strategies, and audience engagement. Our contribution is three-fold: (1) To our knowledge, this is the first attempt to use ad videos with well-designed tasks to evaluate LLMs. We contribute AdsQA, a challenging ad Video QA benchmark derived from 1,544 ad videos with 10,962 clips, totaling 22.7 hours, providing 5 challenging tasks. (2) We propose ReAd-R, a Deepseek-R1 styled RL model that reflects on questions, and generates answers via reward-driven optimization. (3) We benchmark 14 top-tier LLMs on AdsQA, and our \texttt{ReAd-R}~achieves the state-of-the-art outperforming strong competitors equipped with long-chain reasoning capabilities by a clear margin.

Peer review is essential for scientific progress but faces growing challenges due to increasing submission volumes and reviewer fatigue. Existing automated review approaches struggle with factual accuracy, rating consistency, and analytical depth, often generating superficial or generic feedback lacking the insights characteristic of high-quality human reviews. We introduce ReviewRL, a reinforcement learning framework for generating comprehensive and factually grounded scientific paper reviews. Our approach combines: (1) an ArXiv-MCP retrieval-augmented context generation pipeline that incorporates relevant scientific literature, (2) supervised fine-tuning that establishes foundational reviewing capabilities, and (3) a reinforcement learning procedure with a composite reward function that jointly enhances review quality and rating accuracy. Experiments on ICLR 2025 papers demonstrate that ReviewRL significantly outperforms existing methods across both rule-based metrics and model-based quality assessments. ReviewRL establishes a foundational framework for RL-driven automatic critique generation in scientific discovery, demonstrating promising potential for future development in this domain. The implementation of ReviewRL will be released at GitHub.

Alpha-synuclein (α-syn) has long been identified as the etiologic agent of multiple neurodegenerative diseases, the most common and well-known of which are Parkinson’s disease (PD) and Lewy body dementia (LBD). While it is known that the pathophysiology of these synucleinopathies involves aggregation of improperly-folded α-syn, the mechanisms leading to its accumulation have not been fully identified. However, multiple pathways have been proposed, any or all of which may contribute to synucleinopathies. The role of α-syn in normal homeostasis and in other organ systems, especially the hematopoietic system, has been reported recently. Research within the last decade has shown that α-syn plays many vital and conserved roles in the cell biology of various organ systems, such as packaging of cell products, exocytosis, membrane stabilization, and more. This protein has been recognized as an essential factor in normal hematopoietic and immune systems function, and its deficiency leads to an abnormal phenotype, in hematopoietic and immune cell lineages. Similar phenotypes in synucleinopathies not only emphasize the conserved nature of the synuclein family but suggest a bimodal pathophysiology in which aggregated α-syn leads to cellular toxicity while causing derangement of systems that require it. Research into specific molecular mechanisms and potential treatments may provide further understanding of neurodegenerative diseases as well as lead to novel therapies. However, elucidation of the systemic roles of α-syn in addition to its toxicity in excess is essential to prevent treatment-induced deprivation, which paradoxically harms the patient. Here, we address recent advances in systemic synucleinopathies and putative interconnectedness of these compartments. While previous studies and reviews have focused on the mechanisms of α-syn synthesis, transport, and aggregation within systems, this review focuses on the potential inter-systemic nature of synucleinopathies and their possible synergistic origins.

Diffusion models have recently demonstrated exceptional performance in image generation task. However, existing image generation methods still significantly suffer from the dilemma of image reasoning, especially in logic-centered image generation tasks. Inspired by the success of Chain of Thought (CoT) and Reinforcement Learning (RL) in LLMs, we propose SRRL, a self-reflective RL algorithm for diffusion models to achieve reasoning generation of logical images by performing reflection and iteration across generation trajectories. The intermediate samples in the denoising process carry noise, making accurate reward evaluation difficult. To address this challenge, SRRL treats the entire denoising trajectory as a CoT step with multi-round reflective denoising process and introduces condition guided forward process, which allows for reflective iteration between CoT steps. Through SRRL-based iterative diffusion training, we introduce image reasoning through CoT into generation tasks adhering to physical laws and unconventional physical phenomena for the first time. Notably, experimental results of case study exhibit that the superior performance of our SRRL algorithm even compared with GPT-4o. The project page is https://jadenpan0.github.io/srrl.github.io/.

Test-Time Scaling (TTS) is an important method for improving the performance of Large Language Models (LLMs) by using additional computation during the inference phase. However, current studies do not systematically analyze how policy models, Process Reward Models (PRMs), and problem difficulty influence TTS. This lack of analysis limits the understanding and practical use of TTS methods. In this paper, we focus on two core questions: (1) What is the optimal approach to scale test-time computation across different policy models, PRMs, and problem difficulty levels? (2) To what extent can extended computation improve the performance of LLMs on complex tasks, and can smaller language models outperform larger ones through this approach? Through comprehensive experiments on MATH-500 and challenging AIME24 tasks, we have the following observations: (1) The compute-optimal TTS strategy is highly dependent on the choice of policy model, PRM, and problem difficulty. (2) With our compute-optimal TTS strategy, extremely small policy models can outperform larger models. For example, a 1B LLM can exceed a 405B LLM on MATH-500. Moreover, on both MATH-500 and AIME24, a 0.5B LLM outperforms GPT-4o, a 3B LLM surpasses a 405B LLM, and a 7B LLM beats o1 and DeepSeek-R1, while with higher inference efficiency. These findings show the significance of adapting TTS strategies to the specific characteristics of each task and model and indicate that TTS is a promising approach for enhancing the reasoning abilities of LLMs.

Sihang Zeng, Kai Tian, Kaiyan Zhang, Yuru Wang, Junqi Gao, Runze Liu, Sa Yang, Jingxuan Li, Xinwei Long, Jiaheng Ma, Biqing Qi, Bowen Zhou. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025.