Activation-aware Probe-Query: Effective Key-Value Retrieval for Long-Context LLMs Inference

Recent advances in large language models (LLMs) have showcased exceptional performance in long-context tasks, while facing significant inference efficiency challenges with limited GPU memory. Existing solutions first proposed the sliding-window approach to accumulate a set of historical \textbf{key-value} (KV) pairs for reuse, then further improvements selectively retain its subsets at each step. However, due to the sparse attention distribution across a long context, it is hard to identify and recall relevant KV pairs, as the attention is distracted by massive candidate pairs. Additionally, we found it promising to select representative tokens as probe-Query in each sliding window to effectively represent the entire context, which is an approach overlooked by existing methods. Thus, we propose \textbf{ActQKV}, a training-free, \textbf{Act}ivation-aware approach that dynamically determines probe-\textbf{Q}uery and leverages it to retrieve the relevant \textbf{KV} pairs for inference. Specifically, ActQKV monitors a token-level indicator, Activation Bias, within each context window, enabling the proper construction of probe-Query for retrieval at pre-filling stage. To accurately recall the relevant KV pairs and minimize the irrelevant ones, we design a dynamic KV cut-off mechanism guided by information density across layers at the decoding stage. Experiments on the Long-Bench and $\infty$ Benchmarks demonstrate its state-of-the-art performance with competitive inference quality and resource efficiency.

Perovskite-LLM: Knowledge-Enhanced Large Language Models for Perovskite Solar Cell Research

The rapid advancement of perovskite solar cells (PSCs) has led to an exponential growth in research publications, creating an urgent need for efficient knowledge management and reasoning systems in this domain. We present a comprehensive knowledge-enhanced system for PSCs that integrates three key components. First, we develop Perovskite-KG, a domain-specific knowledge graph constructed from 1,517 research papers, containing 23,789 entities and 22,272 relationships. Second, we create two complementary datasets: Perovskite-Chat, comprising 55,101 high-quality question-answer pairs generated through a novel multi-agent framework, and Perovskite-Reasoning, containing 2,217 carefully curated materials science problems. Third, we introduce two specialized large language models: Perovskite-Chat-LLM for domain-specific knowledge assistance and Perovskite-Reasoning-LLM for scientific reasoning tasks. Experimental results demonstrate that our system significantly outperforms existing models in both domain-specific knowledge retrieval and scientific reasoning tasks, providing researchers with effective tools for literature review, experimental design, and complex problem-solving in PSC research.

High-Sensitivity Vision-Based Tactile Sensing Enhanced by Microstructures and Lightweight CNN

Tactile sensing is critical in advanced interactive systems by emulating the human sense of touch to detect stimuli. Vision-based tactile sensors (VBTSs) are promising for their ability to provide rich information, robustness, adaptability, low cost, and multimodal capabilities. However, current technologies still have limitations in sensitivity, spatial resolution, and the high computational demands of deep learning-based image processing. This paper presents a comprehensive approach combining a novel sensor structure with micromachined structures and an efficient image processing method, and demonstrates that carefully engineered microstructures within the sensor hardware can significantly enhance sensitivity while reducing computational load. Unlike traditional designs with tracking markers, our sensor incorporates an interface surface with micromachined trenches, as an example of microstructures, which modulate light transmission and amplify the variation in response to applied force. By capturing variations in brightness, wire width, and cross pattern locations with a camera, the sensor accurately infers the contact location, the magnitude of displacement and applied force with a lightweight convolutional neural network (CNN). Theoretical and experimental results demonstrated that the microstructures significantly enhance sensitivity by amplifying the visual effects of shape distortion. The sensor system effectively detected forces below 10 mN, and achieved a millimetre-level single-point spatial resolution. Using a model with only one convolutional layer, a mean absolute error (MAE) below 0.05 mm have been achieved. Its soft sensor body ensures compatibility with soft robots and wearable electronics, while its immunity to electrical crosstalk and interference guarantees reliability in complex human-machine environments.

Benchmarking Graph Learning for Drug-Drug Interaction Prediction

Predicting drug-drug interaction (DDI) plays an important role in pharmacology and healthcare for identifying potential adverse interactions and beneficial combination therapies between drug pairs. Recently, a flurry of graph learning methods have been introduced to predict drug-drug interactions. However, evaluating existing methods has several limitations, such as the absence of a unified comparison framework for DDI prediction methods, lack of assessments in meaningful real-world scenarios, and insufficient exploration of side information usage. In order to address these unresolved limitations in the literature, we propose a DDI prediction benchmark on graph learning. We first conduct unified evaluation comparison among existing methods. To meet realistic scenarios, we further evaluate the performance of different methods in settings with new drugs involved and examine the performance across different DDI types. Component analysis is conducted on the biomedical network to better utilize side information. Through this work, we hope to provide more insights for the problem of DDI prediction. Our implementation and data is open-sourced at https://anonymous.4open.science/r/DDI-Benchmark-ACD9/.

Knowledge-Aware Parsimony Learning: A Perspective from Relational Graphs

The scaling law, a strategy that involves the brute-force scaling of the training dataset and learnable parameters, has become a prevalent approach for developing stronger learning models. In this paper, we examine its rationale in terms of learning from relational graphs. We demonstrate that directly adhering to such a scaling law does not necessarily yield stronger models due to architectural incompatibility and representation bottlenecks. To tackle this challenge, we propose a novel framework for learning from relational graphs via knowledge-aware parsimony learning. Our method draws inspiration from the duality between data and knowledge inherent in these graphs. Specifically, we first extract knowledge (like symbolic logic and physical laws) during the learning process, and then apply combinatorial generalization to the task at hand. This extracted knowledge serves as the ``building blocks'' for achieving parsimony learning. By applying this philosophy to architecture, parameters, and inference, we can effectively achieve versatile, sample-efficient, and interpretable learning. Experimental results show that our proposed framework surpasses methods that strictly follow the traditional scaling-up roadmap. This highlights the importance of incorporating knowledge in the development of next-generation learning technologies.

R^2AG: Incorporating Retrieval Information into Retrieval Augmented Generation

Retrieval augmented generation (RAG) has been applied in many scenarios to augment large language models (LLMs) with external documents provided by retrievers. However, a semantic gap exists between LLMs and retrievers due to differences in their training objectives and architectures. This misalignment forces LLMs to passively accept the documents provided by the retrievers, leading to incomprehension in the generation process, where the LLMs are burdened with the task of distinguishing these documents using their inherent knowledge. This paper proposes R$^2$AG, a novel enhanced RAG framework to fill this gap by incorporating Retrieval information into Retrieval Augmented Generation. Specifically, R$^2$AG utilizes the nuanced features from the retrievers and employs a R$^2$-Former to capture retrieval information. Then, a retrieval-aware prompting strategy is designed to integrate retrieval information into LLMs' generation. Notably, R$^2$AG suits low-source scenarios where LLMs and retrievers are frozen. Extensive experiments across five datasets validate the effectiveness, robustness, and efficiency of R$^2$AG. Our analysis reveals that retrieval information serves as an anchor to aid LLMs in the generation process, thereby filling the semantic gap.

Explore then Determine: A GNN-LLM Synergy Framework for Reasoning over Knowledge Graph

The task of reasoning over Knowledge Graphs (KGs) poses a significant challenge for Large Language Models (LLMs) due to the complex structure and large amounts of irrelevant information. Existing LLM reasoning methods overlook the importance of compositional learning on KG to supply with precise knowledge. Besides, the fine-tuning and frequent interaction with LLMs incur substantial time and resource costs. This paper focuses on the Question Answering over Knowledge Graph (KGQA) task and proposes an Explore-then-Determine (EtD) framework that synergizes LLMs with graph neural networks (GNNs) for reasoning over KGs. The Explore stage employs a lightweight GNN to explore promising candidates and relevant fine-grained knowledge to the questions, while the Determine stage utilizes the explored information to construct a knowledge-enhanced multiple-choice prompt, guiding a frozen LLM to determine the final answer. Extensive experiments on three benchmark KGQA datasets demonstrate that EtD achieves state-of-the-art performance and generates faithful reasoning results.

BP4ER: Bootstrap Prompting for Explicit Reasoning in Medical Dialogue Generation

Medical dialogue generation (MDG) has gained increasing attention due to its substantial practical value. Previous works typically employ a sequence-to-sequence framework to generate medical responses by modeling dialogue context as sequential text with annotated medical entities. While these methods have been successful in generating fluent responses, they fail to provide process explanations of reasoning and require extensive entity annotation. To address these limitations, we propose the method Bootstrap Prompting for Explicit Reasoning in MDG (BP4ER), which explicitly model MDG's multi-step reasoning process and iteratively enhance this reasoning process. We employ a least-to-most prompting strategy to guide a large language model (LLM) in explicit reasoning, breaking down MDG into simpler sub-questions. These sub-questions build on answers from previous ones. Additionally, we also introduce two distinct bootstrapping techniques for prompting, which autonomously correct errors and facilitate the LLM's explicit reasoning. This approach eliminates the need for entity annotation and increases the transparency of the MDG process by explicitly generating the intermediate reasoning chain. The experimental findings on the two public datasets indicate that BP4ER outperforms state-of-the-art methods in terms of both objective and subjective evaluation metrics.

Knowledge-Enhanced Recommendation with User-Centric Subgraph Network

Recommendation systems, as widely implemented nowadays on various platforms, recommend relevant items to users based on their preferences. The classical methods which rely on user-item interaction matrices has limitations, especially in scenarios where there is a lack of interaction data for new items. Knowledge graph (KG)-based recommendation systems have emerged as a promising solution. However, most KG-based methods adopt node embeddings, which do not provide personalized recommendations for different users and cannot generalize well to the new items. To address these limitations, we propose Knowledge-enhanced User-Centric subgraph Network (KUCNet), a subgraph learning approach with graph neural network (GNN) for effective recommendation. KUCNet constructs a U-I subgraph for each user-item pair that captures both the historical information of user-item interactions and the side information provided in KG. An attention-based GNN is designed to encode the U-I subgraphs for recommendation. Considering efficiency, the pruned user-centric computation graph is further introduced such that multiple U-I subgraphs can be simultaneously computed and that the size can be pruned by Personalized PageRank. Our proposed method achieves accurate, efficient, and interpretable recommendations especially for new items. Experimental results demonstrate the superiority of KUCNet over state-of-the-art KG-based and collaborative filtering (CF)-based methods.

Less is More: One-shot Subgraph Reasoning on Large-scale Knowledge Graphs

To deduce new facts on a knowledge graph (KG), a link predictor learns from the graph structure and collects local evidence to find the answer to a given query. However, existing methods suffer from a severe scalability problem due to the utilization of the whole KG for prediction, which hinders their promise on large scale KGs and cannot be directly addressed by vanilla sampling methods. In this work, we propose the one-shot-subgraph link prediction to achieve efficient and adaptive prediction. The design principle is that, instead of directly acting on the whole KG, the prediction procedure is decoupled into two steps, i.e., (i) extracting only one subgraph according to the query and (ii) predicting on this single, query dependent subgraph. We reveal that the non-parametric and computation-efficient heuristics Personalized PageRank (PPR) can effectively identify the potential answers and supporting evidence. With efficient subgraph-based prediction, we further introduce the automated searching of the optimal configurations in both data and model spaces. Empirically, we achieve promoted efficiency and leading performances on five large-scale benchmarks. The code is publicly available at: https://github.com/tmlr-group/one-shot-subgraph.