Recent advancements in reasoning with large language models (RLLMs), such as OpenAI-O1 and DeepSeek-R1, have demonstrated their impressive capabilities in complex domains like mathematics and coding. A central factor in their success lies in the application of long chain-of-thought (Long CoT) characteristics, which enhance reasoning abilities and enable the solution of intricate problems. However, despite these developments, a comprehensive survey on Long CoT is still lacking, limiting our understanding of its distinctions from traditional short chain-of-thought (Short CoT) and complicating ongoing debates on issues like "overthinking" and "inference-time scaling." This survey seeks to fill this gap by offering a unified perspective on Long CoT. (1) We first distinguish Long CoT from Short CoT and introduce a novel taxonomy to categorize current reasoning paradigms. (2) Next, we explore the key characteristics of Long CoT: deep reasoning, extensive exploration, and feasible reflection, which enable models to handle more complex tasks and produce more efficient, coherent outcomes compared to the shallower Short CoT. (3) We then investigate key phenomena such as the emergence of Long CoT with these characteristics, including overthinking, and inference-time scaling, offering insights into how these processes manifest in practice. (4) Finally, we identify significant research gaps and highlight promising future directions, including the integration of multi-modal reasoning, efficiency improvements, and enhanced knowledge frameworks. By providing a structured overview, this survey aims to inspire future research and further the development of logical reasoning in artificial intelligence.

Abstract Despite rapid progress in AI agents for computer-aided drug design (CADD), protein-ligand simulation workflows remain fragmented across disparate tools, creating a major bottleneck for scalable candidate evaluation. Here, we present PRISM ( P rotein- R eceptor I nteraction Simulation M odeler), a Python platform built on GROMACS that unifies ligand parameterization across multiple force fields, automated system construction, enhanced sampling, multi-tier binding free energy estimation, and trajectory analysis within a single workflow. Through the Model Context Protocol (MCP), PRISM further serves as the computational infrastructure for CADD-Agent , an expert-workflow-driven AI agent designed to orchestrate hierarchical drug screening pipelines. As a pilot application, we applied PRISM to riboflavin synthase and demonstrated end-to-end automation from candidate library assembly to binding pocket characterization, identifying a potential allosteric inhibition site at the oligomerization interface. Together, these results establish PRISM as a high-throughput simulation infrastructure for agent-enabled CADD.

Unstructured model editing aims to update models with real-world text, yet existing methods often memorize text holistically without reliable fine-grained fact access. To address this, we propose FABLE, a hierarchical framework that decouples fine-grained fact injection from holistic text generation. FABLE follows a two-stage, fact-first strategy: discrete facts are anchored in shallow layers, followed by minimal updates to deeper layers to produce coherent text. This decoupling resolves the mismatch between holistic recall and fine-grained fact access, reflecting the unidirectional Transformer flow in which surface-form generation amplifies rather than corrects underlying fact representations. We also introduce UnFine, a diagnostic benchmark with fine-grained question-answer pairs and fact-level metrics for systematic evaluation. Experiments show that FABLE substantially improves fine-grained question answering while maintaining state-of-the-art holistic editing performance. Our code is publicly available at https://github.com/caskcsg/FABLE.

When using supervised fine-tuning (SFT) to adapt large language models (LLMs) to specific domains, a significant challenge arises: should we use the entire SFT dataset for fine-tuning? Common practice often involves fine-tuning directly on the entire dataset due to limited information on the LLM's past training data. However, if the SFT dataset largely overlaps with the model's existing knowledge, the performance gains are minimal, leading to wasted computational resources. Identifying the unknown knowledge within the SFT dataset and using it to fine-tune the model could substantially improve the training efficiency. To address this challenge, we propose a self-learning framework for LLMs inspired by human learning pattern. This framework takes a fine-tuning (SFT) dataset in a specific domain as input. First, the LLMs answer the questions in the SFT dataset. The LLMs then objectively grade the responses and filter out the incorrectly answered QA pairs. Finally, we fine-tune the LLMs based on this filtered QA set. Experimental results in the fields of agriculture and medicine demonstrate that our method substantially reduces training time while achieving comparable improvements to those attained with full dataset fine-tuning. By concentrating on the unknown knowledge within the SFT dataset, our approach enhances the efficiency of fine-tuning LLMs.

Acoustic Leak Detection (ALD) plays a pivotal role in ensuring the operational safety of water distribution networks (WDNs). However, the cross-domain deployment of single-scenario ALD models is significantly hindered by environmental heterogeneity (pipe materials and diameters, etc.) and data scarcity in practical WDNs. This study presents the first systematic investigation into the generalizability and transferability of ALD models across multi-source WDNs through comprehensive cross-domain evaluation. The results show that: (1) The global model trained on the multi-region WDNs exhibits better generalization ability with an average accuracy improvement of about 2 % compared to the local model. (2) Fine Tuning strategy achieves high transfer performance (96.8 % and 96.2 % accuracy for cross-material and cross-diameter scenarios respectively), outperforming Direct Transfer and Feature Extraction methods. (3) Transfer asymmetry is related to the distribution of input data under different conditions, where metal-to-nonmetal and large-to-small diameter transfers exhibit enhanced adaptability through broader source-domain frequency coverage. (4) Target-domain data requirements with the Fine Tuning strategy can be reduced by 50 % while maintaining superior accuracy compared to source-domain local models. These findings advance AI-driven ALD techniques from a single-scenario-specific perspective to more mature applications of multi-scenario-universal.

We present a generative dialogue system capable of operating in a full-duplex manner, allowing for seamless interaction. It is based on a large language model (LLM) carefully aligned to be aware of a perception module, a motor function module, and the concept of a simple finite state machine (called neural FSM) with two states. The perception and motor function modules operate in tandem, allowing the system to speak and listen to the user simultaneously. The LLM generates textual tokens for inquiry responses and makes autonomous decisions to start responding to, wait for, or interrupt the user by emitting control tokens to the neural FSM. All these tasks of the LLM are carried out as next token prediction on a serialized view of the dialogue in real-time. In automatic quality evaluations simulating real-life interaction, the proposed system reduces the average conversation response latency by more than threefold compared with LLM-based half-duplex dialogue systems while responding within less than 500 milliseconds in more than 50% of evaluated interactions. Running an LLM with only 8 billion parameters, our system exhibits an 8% higher interruption precision rate than the best available commercial LLM for voice-based dialogue.

Previous knowledge graph embedding approaches usually map entities to representations and utilize score functions to predict the target entities, yet they typically struggle to reason rare or emerging unseen entities. In this paper, we propose kNN-KGE, a new knowledge graph embedding approach with pre-trained language models, by linearly interpolating its entity distribution with k-nearest neighbors. We compute the nearest neighbors based on the distance in the entity embedding space from the knowledge store. Our approach can allow rare or emerging entities to be memorized explicitly rather than implicitly in model parameters. Experimental results demonstrate that our approach can improve inductive and transductive link prediction results and yield better performance for low-resource settings with only a few triples, which might be easier to reason via explicit memory (Code is available at: https://github.com/zjunlp/KNN-KG ).