Shufei Zhang

Junxian Li, Di Zhang, Xunzhi Wang et al. · mit

Large Language Models (LLMs) have achieved remarkable success and have been applied across various scientific fields, including chemistry. However, many chemical tasks require the processing of visual information, which cannot be successfully handled by existing chemical LLMs. This brings a growing need for models capable of integrating multimodal information in the chemical domain. In this paper, we introduce \textbf{ChemVLM}, an open-source chemical multimodal large language model specifically designed for chemical applications. ChemVLM is trained on a carefully curated bilingual multimodal dataset that enhances its ability to understand both textual and visual chemical information, including molecular structures, reactions, and chemistry examination questions. We develop three datasets for comprehensive evaluation, tailored to Chemical Optical Character Recognition (OCR), Multimodal Chemical Reasoning (MMCR), and Multimodal Molecule Understanding tasks. We benchmark ChemVLM against a range of open-source and proprietary multimodal large language models on various tasks. Experimental results demonstrate that ChemVLM achieves competitive performance across all evaluated tasks. Our model can be found at https://huggingface.co/AI4Chem/ChemVLM-26B.

Yiheng Wang, Yixin Chen, Shuo Li et al.

We introduce SciEvalKit, a unified benchmarking toolkit designed to evaluate AI models for science across a broad range of scientific disciplines and task capabilities. Unlike general-purpose evaluation platforms, SciEvalKit focuses on the core competencies of scientific intelligence, including Scientific Multimodal Perception, Scientific Multimodal Reasoning, Scientific Multimodal Understanding, Scientific Symbolic Reasoning, Scientific Code Generation, Science Hypothesis Generation and Scientific Knowledge Understanding. It supports six major scientific domains, spanning from physics and chemistry to astronomy and materials science. SciEvalKit builds a foundation of expert-grade scientific benchmarks, curated from real-world, domain-specific datasets, ensuring that tasks reflect authentic scientific challenges. The toolkit features a flexible, extensible evaluation pipeline that enables batch evaluation across models and datasets, supports custom model and dataset integration, and provides transparent, reproducible, and comparable results. By bridging capability-based evaluation and disciplinary diversity, SciEvalKit offers a standardized yet customizable infrastructure to benchmark the next generation of scientific foundation models and intelligent agents. The toolkit is open-sourced and actively maintained to foster community-driven development and progress in AI4Science.

Wanghan Xu, Yuhao Zhou, Yifan Zhou et al.

Despite advances in scientific AI, a coherent framework for Scientific General Intelligence (SGI)-the ability to autonomously conceive, investigate, and reason across scientific domains-remains lacking. We present an operational SGI definition grounded in the Practical Inquiry Model (PIM: Deliberation, Conception, Action, Perception) and operationalize it via four scientist-aligned tasks: deep research, idea generation, dry/wet experiments, and experimental reasoning. SGI-Bench comprises over 1,000 expert-curated, cross-disciplinary samples inspired by Science's 125 Big Questions, enabling systematic evaluation of state-of-the-art LLMs. Results reveal gaps: low exact match (10--20%) in deep research despite step-level alignment; ideas lacking feasibility and detail; high code executability but low execution result accuracy in dry experiments; low sequence fidelity in wet protocols; and persistent multimodal comparative-reasoning challenges. We further introduce Test-Time Reinforcement Learning (TTRL), which optimizes retrieval-augmented novelty rewards at inference, enhancing hypothesis novelty without reference answer. Together, our PIM-grounded definition, workflow-centric benchmark, and empirical insights establish a foundation for AI systems that genuinely participate in scientific discovery.

Zeyu Xia, Jinzhe Ma, Congjie Zheng et al.

Large Language Models (LLMs) have emerged as powerful tools for accelerating scientific discovery, yet their static knowledge and hallucination issues hinder autonomous research applications. Recent advances integrate LLMs into agentic frameworks, enabling retrieval, reasoning, and tool use for complex scientific workflows. Here, we present a domain-specialized agent designed for reliable automation of first-principles materials computations. By embedding domain expertise, the agent ensures physically coherent multi-step workflows and consistently selects convergent, well-posed parameters, thereby enabling reliable end-to-end computational execution. A new benchmark of diverse computational tasks demonstrates that our system significantly outperforms standalone LLMs in both accuracy and robustness. This work establishes a verifiable foundation for autonomous computational experimentation and represents a key step toward fully automated scientific discovery.

Weida Wang, Dongchen Huang, Jiatong Li et al.

We introduce CMPhysBench, designed to assess the proficiency of Large Language Models (LLMs) in Condensed Matter Physics, as a novel Benchmark. CMPhysBench is composed of more than 520 graduate-level meticulously curated questions covering both representative subfields and foundational theoretical frameworks of condensed matter physics, such as magnetism, superconductivity, strongly correlated systems, etc. To ensure a deep understanding of the problem-solving process,we focus exclusively on calculation problems, requiring LLMs to independently generate comprehensive solutions. Meanwhile, leveraging tree-based representations of expressions, we introduce the Scalable Expression Edit Distance (SEED) score, which provides fine-grained (non-binary) partial credit and yields a more accurate assessment of similarity between prediction and ground-truth. Our results show that even the best models, Grok-4, reach only 36 average SEED score and 28% accuracy on CMPhysBench, underscoring a significant capability gap, especially for this practical and frontier domain relative to traditional physics. The code anddataset are publicly available at https://github.com/CMPhysBench/CMPhysBench.

Jiaqing Xie, Weida Wang, Ben Gao et al.

Quantitative chemistry is central to modern chemical research, yet the ability of large language models (LLMs) to perform its rigorous, step-by-step calculations remains underexplored. To fill this blank, we propose QCBench, a Quantitative Chemistry oriented benchmark comprising 350 computational chemistry problems across 7 chemistry subfields, which contains analytical chemistry, bio/organic chemistry, general chemistry, inorganic chemistry, physical chemistry, polymer chemistry and quantum chemistry. To systematically evaluate the mathematical reasoning abilities of large language models (LLMs), they are categorized into three tiers: easy, medium, and difficult. Each problem, rooted in realistic chemical scenarios, is structured to prevent heuristic shortcuts and demand explicit numerical reasoning. QCBench enables fine-grained diagnosis of computational weaknesses, reveals model-specific limitations across difficulty levels, and lays the groundwork for future improvements such as domain-adaptive fine-tuning or multi-modal integration. Evaluations on 24 LLMs demonstrate a consistent performance degradation with increasing task complexity, highlighting the current gap between language fluency and scientific computation accuracy. Code for QCBench is available at https://github.com/jiaqingxie/QCBench.

Haonan He, Yuchen Ren, Yining Tang et al.

Large language models (LLMs) have shown remarkable capabilities in general domains, but their application to multi-omics biology remains underexplored. To address this gap, we introduce Biology-Instructions, the first large-scale instruction-tuning dataset for multi-omics biological sequences, including DNA, RNA, proteins, and multi-molecules. This dataset bridges LLMs and complex biological sequence-related tasks, enhancing their versatility and reasoning while maintaining conversational fluency. We also highlight significant limitations of current state-of-the-art LLMs on multi-omics tasks without specialized training. To overcome this, we propose ChatMultiOmics, a strong baseline with a novel three-stage training pipeline, demonstrating superior biological understanding through Biology-Instructions. Both resources are publicly available, paving the way for better integration of LLMs in multi-omics analysis. The Biology-Instructions is publicly available at: https://github.com/hhnqqq/Biology-Instructions.

Zhuang Qian, Shufei Zhang, Kaizhu Huang et al.

In this work, we consider model robustness of deep neural networks against adversarial attacks from a global manifold perspective. Leveraging both the local and global latent information, we propose a novel adversarial training method through robust optimization, and a tractable way to generate Latent Manifold Adversarial Examples (LMAEs) via an adversarial game between a discriminator and a classifier. The proposed adversarial training with latent distribution (ATLD) method defends against adversarial attacks by crafting LMAEs with the latent manifold in an unsupervised manner. ATLD preserves the local and global information of latent manifold and promises improved robustness against adversarial attacks. To verify the effectiveness of our proposed method, we conduct extensive experiments over different datasets (e.g., CIFAR-10, CIFAR-100, SVHN) with different adversarial attacks (e.g., PGD, CW), and show that our method substantially outperforms the state-of-the-art (e.g., Feature Scattering) in adversarial robustness by a large accuracy margin. The source codes are available at https://github.com/LitterQ/ATLD-pytorch.

Di Zhang, Wei Liu, Qian Tan et al.

Large language models (LLMs) have made impressive progress in chemistry applications. However, the community lacks an LLM specifically designed for chemistry. The main challenges are two-fold: firstly, most chemical data and scientific knowledge are stored in structured databases, which limits the model's ability to sustain coherent dialogue when used directly. Secondly, there is an absence of objective and fair benchmark that encompass most chemistry tasks. Here, we introduce ChemLLM, a comprehensive framework that features the first LLM dedicated to chemistry. It also includes ChemData, a dataset specifically designed for instruction tuning, and ChemBench, a robust benchmark covering nine essential chemistry tasks. ChemLLM is adept at performing various tasks across chemical disciplines with fluid dialogue interaction. Notably, ChemLLM achieves results comparable to GPT-4 on the core chemical tasks and demonstrates competitive performance with LLMs of similar size in general scenarios. ChemLLM paves a new path for exploration in chemical studies, and our method of incorporating structured chemical knowledge into dialogue systems sets a new standard for developing LLMs in various scientific fields. Codes, Datasets, and Model weights are publicly accessible at https://hf.co/AI4Chem

InternAgent Team, Bo Zhang, Shiyang Feng et al.

Artificial Intelligence (AI) is accelerating the transformation of scientific research paradigms, not only enhancing research efficiency but also driving innovation. We introduce InternAgent, a unified closed-loop multi-agent framework to conduct Autonomous Scientific Research (ASR) across various scientific research fields, enabling researchers to tackle complicated problems in these fields with unprecedented speed and precision. InternAgent highlights three key advantages: 1) Scalability: InternAgent has demonstrated its versatility across 12 scientific research tasks, capable of generating innovative ideas to enhance the performance of baseline code. 2) Interactivity: InternAgent provides an interface for human expert feedback and multi-agent interaction in automated end-to-end processes, allowing for the seamless integration of domain expert knowledge. 3) Efficiency: InternAgent has achieved promising performance gains in several scientific fields with significantly less time cost compared to human efforts. For instance, in reaction yield prediction, it increased from 27.6% to 35.4% in just 12 hours; in enhancer activity prediction, accuracy rose from 0.65 to 0.79 with only 4 hours of processing; and in 2D semantic segmentation, precision advanced from 78.8% to 81.0% in a mere 30 hours.

Zhihao Dou, Qinjian Zhao, Zhongwei Wan et al.

Large language models (LLMs) have demonstrated remarkable reasoning abilities in complex tasks, often relying on Chain-of-Thought (CoT) reasoning. However, due to their autoregressive token-level generation, the reasoning process is largely constrained to local decision-making and lacks global planning. This limitation frequently results in redundant, incoherent, or inaccurate reasoning, which significantly degrades overall performance. Existing approaches, such as tree-based algorithms and reinforcement learning (RL), attempt to address this issue but suffer from high computational costs and often fail to produce optimal reasoning trajectories. To tackle this challenge, we propose Plan-Then-Action Enhanced Reasoning with Group Relative Policy Optimization PTA-GRPO, a two-stage framework designed to improve both high-level planning and fine-grained CoT reasoning. In the first stage, we leverage advanced LLMs to distill CoT into compact high-level guidance, which is then used for supervised fine-tuning (SFT). In the second stage, we introduce a guidance-aware RL method that jointly optimizes the final output and the quality of high-level guidance, thereby enhancing reasoning effectiveness. We conduct extensive experiments on multiple mathematical reasoning benchmarks, including MATH, AIME2024, AIME2025, and AMC, across diverse base models such as Qwen2.5-7B-Instruct, Qwen3-8B, Qwen3-14B, and LLaMA3.2-3B. Experimental results demonstrate that PTA-GRPO consistently achieves stable and significant improvements across different models and tasks, validating its effectiveness and generalization.

Jiatong Li, Weida Wang, Qinggang Zhang et al.

Large language models (LLMs), especially Explicit Long Chain-of-Thought (CoT) reasoning models like DeepSeek-R1 and QWQ, have demonstrated powerful reasoning capabilities, achieving impressive performance in commonsense reasoning and mathematical inference. Despite their effectiveness, Long-CoT reasoning models are often criticized for their limited ability and low efficiency in knowledge-intensive domains such as molecule discovery. Success in this field requires a precise understanding of domain knowledge, including molecular structures and chemical principles, which is challenging due to the inherent complexity of molecular data and the scarcity of high-quality expert annotations. To bridge this gap, we introduce Mol-R1, a novel framework designed to improve explainability and reasoning performance of R1-like Explicit Long-CoT reasoning LLMs in text-based molecule generation. Our approach begins with a high-quality reasoning dataset curated through Prior Regulation via In-context Distillation (PRID), a dedicated distillation strategy to effectively generate paired reasoning traces guided by prior regulations. Building upon this, we introduce MoIA, Molecular Iterative Adaptation, a sophisticated training strategy that iteratively combines Supervised Fine-tuning (SFT) with Reinforced Policy Optimization (RPO), tailored to boost the reasoning performance of R1-like reasoning models for molecule discovery. Finally, we examine the performance of Mol-R1 in the text-based molecule reasoning generation task, showing superior performance against existing baselines.

Ruifeng Li, Dongzhan Zhou, Ancheng Shen et al.

Artificial intelligence (AI) technology has demonstrated remarkable potential in drug dis-covery, where pharmacokinetics plays a crucial role in determining the dosage, safety, and efficacy of new drugs. A major challenge for AI-driven drug discovery (AIDD) is the scarcity of high-quality data, which often requires extensive wet-lab work. A typical example of this is pharmacokinetic experiments. In this work, we develop a physical formula enhanced mul-ti-task learning (PEMAL) method that predicts four key parameters of pharmacokinetics simultaneously. By incorporating physical formulas into the multi-task framework, PEMAL facilitates effective knowledge sharing and target alignment among the pharmacokinetic parameters, thereby enhancing the accuracy of prediction. Our experiments reveal that PEMAL significantly lowers the data demand, compared to typical Graph Neural Networks. Moreover, we demonstrate that PEMAL enhances the robustness to noise, an advantage that conventional Neural Networks do not possess. Another advantage of PEMAL is its high flexibility, which can be potentially applied to other multi-task machine learning scenarios. Overall, our work illustrates the benefits and potential of using PEMAL in AIDD and other scenarios with data scarcity and noise.

Dong Han, Zhehong Ai, Pengxiang Cai et al.

Bayesian optimization (BO) is a powerful tool for scientific discovery in chemistry, yet its efficiency is often hampered by the sparse experimental data and vast search space. Here, we introduce ChemBOMAS: a large language model (LLM)-enhanced multi-agent system that accelerates BO through synergistic data- and knowledge-driven strategies. Firstly, the data-driven strategy involves an 8B-scale LLM regressor fine-tuned on a mere 1% labeled samples for pseudo-data generation, robustly initializing the optimization process. Secondly, the knowledge-driven strategy employs a hybrid Retrieval-Augmented Generation approach to guide LLM in dividing the search space while mitigating LLM hallucinations. An Upper Confidence Bound algorithm then identifies high-potential subspaces within this established partition. Across the LLM-refined subspaces and supported by LLM-generated data, BO achieves the improvement of effectiveness and efficiency. Comprehensive evaluations across multiple scientific benchmarks demonstrate that ChemBOMAS set a new state-of-the-art, accelerating optimization efficiency by up to 5-fold compared to baseline methods.

Di Zhang, Weida Wang, Junxian Li et al.

This paper target in addressing the challenges of underthinking and overthinking in long chain-of-thought (CoT) reasoning for Large Reasoning Models (LRMs) by introducing Reasoning Control Fields (RCF)--a novel test-time approach that injects structured control signals to guide reasoning from a tree search perspective. RCF enables models to adjust reasoning effort according to given control conditions when solving complex tasks. Additionally, we present the Control-R-4K dataset, which consists of challenging problems annotated with detailed reasoning processes and corresponding control fields. To further enhance reasoning control, we propose a Conditional Distillation Finetuning (CDF) method, which trains model--particularly Control-R-32B--to effectively adjust reasoning effort during test time. Experimental results on benchmarks such as AIME2024 and MATH500 demonstrate that our approach achieves state-of-the-art performance at the 32B scale while enabling a controllable Long CoT reasoning process (L-CoT). Overall, this work introduces an effective paradigm for controllable test-time scaling reasoning.

Chenru Jiang, Kaizhu Huang, Shufei Zhang et al.

In this paper, we focus on tackling the precise keypoint coordinates regression task. Most existing approaches adopt complicated networks with a large number of parameters, leading to a heavy model with poor cost-effectiveness in practice. To overcome this limitation, we develop a small yet discrimicative model called STair Network, which can be simply stacked towards an accurate multi-stage pose estimation system. Specifically, to reduce computational cost, STair Network is composed of novel basic feature extraction blocks which focus on promoting feature diversity and obtaining rich local representations with fewer parameters, enabling a satisfactory balance on efficiency and performance. To further improve the performance, we introduce two mechanisms with negligible computational cost, focusing on feature fusion and replenish. We demonstrate the effectiveness of the STair Network on two standard datasets, e.g., 1-stage STair Network achieves a higher accuracy than HRNet by 5.5% on COCO test dataset with 80\% fewer parameters and 68% fewer GFLOPs.

Shufei Zhang, Zhuang Qian, Kaizhu Huang et al.

Generative adversarial networks (GANs) are powerful generative models, but usually suffer from instability and generalization problem which may lead to poor generations. Most existing works focus on stabilizing the training of the discriminator while ignoring the generalization properties. In this work, we aim to improve the generalization capability of GANs by promoting the local robustness within the small neighborhood of the training samples. We also prove that the robustness in small neighborhood of training sets can lead to better generalization. Particularly, we design a robust optimization framework where the generator and discriminator compete with each other in a \textit{worst-case} setting within a small Wasserstein ball. The generator tries to map \textit{the worst input distribution} (rather than a Gaussian distribution used in most GANs) to the real data distribution, while the discriminator attempts to distinguish the real and fake distribution \textit{with the worst perturbation}. We have proved that our robust method can obtain a tighter generalization upper bound than traditional GANs under mild assumptions, ensuring a theoretical superiority of RGAN over GANs. A series of experiments on CIFAR-10, STL-10 and CelebA datasets indicate that our proposed robust framework can improve on five baseline GAN models substantially and consistently.

Shufei Zhang, Kaizhu Huang, Zenglin Xu

We study the model robustness against adversarial examples, referred to as small perturbed input data that may however fool many state-of-the-art deep learning models. Unlike previous research, we establish a novel theory addressing the robustness issue from the perspective of stability of the loss function in the small neighborhood of natural examples. We propose to exploit an energy function to describe the stability and prove that reducing such energy guarantees the robustness against adversarial examples. We also show that the traditional training methods including adversarial training with the $l_2$ norm constraint (AT) and Virtual Adversarial Training (VAT) tend to minimize the lower bound of our proposed energy function. We make an analysis showing that minimization of such lower bound can however lead to insufficient robustness within the neighborhood around the input sample. Furthermore, we design a more rational method with the energy regularization which proves to achieve better robustness than previous methods. Through a series of experiments, we demonstrate the superiority of our model on both supervised tasks and semi-supervised tasks. In particular, our proposed adversarial framework achieves the best performance compared with previous adversarial training methods on benchmark datasets MNIST, CIFAR-10, and SVHN. Importantly, they demonstrate much better robustness against adversarial examples than all the other comparison methods.

Shufei Zhang, Kaizhu Huang, Jianke Zhu et al.

Recently proposed adversarial training methods show the robustness to both adversarial and original examples and achieve state-of-the-art results in supervised and semi-supervised learning. All the existing adversarial training methods consider only how the worst perturbed examples (i.e., adversarial examples) could affect the model output. Despite their success, we argue that such setting may be in lack of generalization, since the output space (or label space) is apparently less informative.In this paper, we propose a novel method, called Manifold Adversarial Training (MAT). MAT manages to build an adversarial framework based on how the worst perturbation could affect the distributional manifold rather than the output space. Particularly, a latent data space with the Gaussian Mixture Model (GMM) will be first derived.On one hand, MAT tries to perturb the input samples in the way that would rough the distributional manifold the worst. On the other hand, the deep learning model is trained trying to promote in the latent space the manifold smoothness, measured by the variation of Gaussian mixtures (given the local perturbation around the data point). Importantly, since the latent space is more informative than the output space, the proposed MAT can learn better a robust and compact data representation, leading to further performance improvement. The proposed MAT is important in that it can be considered as a superset of one recently-proposed discriminative feature learning approach called center loss. We conducted a series of experiments in both supervised and semi-supervised learning on three benchmark data sets, showing that the proposed MAT can achieve remarkable performance, much better than those of the state-of-the-art adversarial approaches. We also present a series of visualization which could generate further understanding or explanation on adversarial examples.