Guichao Zhu, Lintian Lei, Yuhao Qing, Yichao Fu, Fanxin Li, Dong Huang, Zekai Sun, Heming Cui. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Recent advancements in large language models (LLMs) have revolutionized research discovery across various scientific disciplines, including materials science. The discovery of novel materials, particularly crystal materials, is essential for achieving sustainable development goals (SDGs), as they drive breakthroughs in climate change mitigation, clean and affordable energy, and the promotion of industrial innovation. However, unlocking the full potential of LLMs in materials research remains challenging due to the lack of high-quality, diverse, and instruction-based datasets. Such datasets are crucial for guiding these models in understanding and predicting the structure, property, and function of materials across various tasks. To address this limitation, we introduce Mat-Instruction, a large-scale inorganic material instruction dataset, specifically designed to unlock the potential of LLMs in materials science. Extensive experiments on fine-tuning LLaMA with our Mat-Instruction dataset demonstrate its effectiveness in advancing progress for materials science. The code and dataset are available at https://github.com/zjuKeLiu/Mat-Instructions

Matrix factorization (MF), a cornerstone of recommender systems, decomposes user-item interaction matrices into latent representations. Traditional MF approaches, however, employ a two-stage, non-end-to-end paradigm, sequentially performing recommendation and clustering, resulting in prohibitive computational costs for large-scale applications like e-commerce and IoT, where billions of users interact with trillions of items. To address this, we propose Matrix Factorization with Dynamic Multi-view Clustering (MFDMC), a unified framework that balances efficient end-to-end training with comprehensive utilization of web-scale data and enhances interpretability. MFDMC leverages dynamic multi-view clustering to learn user and item representations, adaptively pruning poorly formed clusters. Each entity's representation is modeled as a weighted projection of robust clusters, capturing its diverse roles across views. This design maximizes representation space utilization, improves interpretability, and ensures resilience for downstream tasks. Extensive experiments demonstrate MFDMC's superior performance in recommender systems and other representation learning domains, such as computer vision, highlighting its scalability and versatility.

Recent advancements in offline reinforcement learning (RL) have underscored the capabilities of conditional sequence modeling (CSM), a paradigm that models the action distribution conditioned on both historical trajectories and target returns associated with each state. However, due to the imbalanced return distribution caused by suboptimal datasets, CSM is grappling with a serious distributional shift problem when conditioning on high returns. While recent approaches attempt to empirically tackle this challenge through return rebalancing techniques such as weighted sampling and value-regularized supervision, the relationship between return rebalancing and the performance of CSM methods is not well understood. In this paper, we reveal that both expert-level and full-spectrum return-coverage critically influence the performance and sample efficiency of CSM policies. Building on this finding, we devise a simple yet effective return-coverage rebalancing mechanism that can be seamlessly integrated into common CSM frameworks, including the most widely used one, Decision Transformer (DT). The resulting CSM algorithm, referred to as Return-rebalanced Value-regularized Decision Transformer (RVDT), integrates both implicit and explicit return-coverage rebalancing mechanisms, and achieves state-of-the-art performance in the D4RL experiments.

Test-time reasoning algorithms such as chain-of-thought, self-consistency, and MCTS enhance LLM problem-solving but can wastefully generate many tokens without improving accuracy. At the same time, we observe that these algorithms exhibit answer stabilization: their intermediate solutions often cease to change after a certain point, and further investment of compute does not change their final answer. To quantify this phenomenon, we introduce Certaindex, an algorithm-agnostic metric measuring this evolving stability, signaling when further computation is unlikely to alter the final result. Certaindex is lightweight, can accelerate reasoning program inference via early exit, and further enables dynamic token allocation, gang scheduling, and many opportunities when integrated with real-world LLM serving systems. To quantify real-world benefits, we built Certaindex as a scheduler into Dynasor, our reasoning-aware LLM serving system, and demonstrate up to 50\% compute savings and 3.3$\times$ higher throughput in real workloads with no accuracy drop. Our code is available at https://github.com/hao-ai-lab/Dynasor.git

Reasoning models excel by generating long chain-of-thoughts, but decoding the resulting thousands of tokens is slow. Token-level specualtive decoding (SD) helps, but its benefit is capped, because the chance that an entire $\gamma$-token guess is correct falls exponentially as $\gamma$ grows. This means allocating more compute for longer token drafts faces an algorithmic ceiling -- making the speedup modest and hardware-agnostic. We raise this ceiling with lookahead reasoning, which exploits a second, step-level layer of parallelism. Our key insight is that reasoning models generate step-by-step, and each step needs only to be semantically correct, not exact token matching. In lookahead reasoning, a lightweight draft model proposes several future steps; the target model expands each proposal in one batched pass, and a verifier keeps semantically correct steps while letting the target regenerate any that fail. Token-level SD still operates within each reasoning step, so the two layers of parallelism multiply. We show lookahead reasoning lifts the peak speedup of SD both theoretically and empirically. Across GSM8K, AIME, and other benchmarks, lookahead reasoning improves the speedup of SD from 1.4x to 2.1x while preserving answer quality, and its speedup scales better with additional GPU throughput. Our code is available at https://github.com/hao-ai-lab/LookaheadReasoning

Retinal vessel segmentation is critical for medical diagnosis, yet existing models often struggle to generalize across domains due to appearance variability, limited annotations, and complex vascular morphology. We propose GraphSeg, a variational Bayesian framework that integrates anatomical graph priors with structure-aware image decomposition to enhance cross-domain segmentation. GraphSeg factorizes retinal images into structure-preserved and structure-degraded components, enabling domain-invariant representation. A deformable graph prior, derived from a statistical retinal atlas, is incorporated via a differentiable alignment and guided by an unsupervised energy function. Experiments on three public benchmarks (CHASE, DRIVE, HRF) show that GraphSeg consistently outperforms existing methods under domain shifts. These results highlight the importance of jointly modeling anatomical topology and image structure for robust generalizable vessel segmentation.

Autoregressive decoding of large language models (LLMs) is memory bandwidth bounded, resulting in high latency and significant wastes of the parallel processing power of modern accelerators. Existing methods for accelerating LLM decoding often require a draft model (e.g., speculative decoding), which is nontrivial to obtain and unable to generalize. In this paper, we introduce Lookahead decoding, an exact, parallel decoding algorithm that accelerates LLM decoding without needing auxiliary models or data stores. It allows trading per-step log(FLOPs) to reduce the number of total decoding steps, is more parallelizable on single or multiple modern accelerators, and is compatible with concurrent memory-efficient attention (e.g., FlashAttention). Our implementation of Lookahead decoding can speed up autoregressive decoding by up to 1.8x on MT-bench and 4x with strong scaling on multiple GPUs in code completion tasks. Our code is avialable at https://github.com/hao-ai-lab/LookaheadDecoding

Autoregressive Large Language Models (LLMs) have achieved impressive performance in language tasks but face two significant bottlenecks: (1) quadratic complexity in the attention module as the number of tokens increases, and (2) limited efficiency due to the sequential processing nature of autoregressive LLMs during generation. While linear attention and speculative decoding offer potential solutions, their applicability and synergistic potential for enhancing autoregressive LLMs remain uncertain. We conduct the first comprehensive study on the efficacy of existing linear attention methods for autoregressive LLMs, integrating them with speculative decoding. We introduce an augmentation technique for linear attention that ensures compatibility with speculative decoding, enabling more efficient training and serving of LLMs. Extensive experiments and ablation studies involving seven existing linear attention models and five encoder/decoder-based LLMs consistently validate the effectiveness of our augmented linearized LLMs. Notably, our approach achieves up to a 6.67 reduction in perplexity on the LLaMA model and up to a 2$\times$ speedup during generation compared to prior linear attention methods. Codes and models are available at https://github.com/GATECH-EIC/Linearized-LLM.

In Large Language Model (LLM) inference, the output length of an LLM request is typically regarded as not known a priori. Consequently, most LLM serving systems employ a simple First-come-first-serve (FCFS) scheduling strategy, leading to Head-Of-Line (HOL) blocking and reduced throughput and service quality. In this paper, we reexamine this assumption -- we show that, although predicting the exact generation length of each request is infeasible, it is possible to predict the relative ranks of output lengths in a batch of requests, using learning to rank. The ranking information offers valuable guidance for scheduling requests. Building on this insight, we develop a novel scheduler for LLM inference and serving that can approximate the shortest-job-first (SJF) schedule better than existing approaches. We integrate this scheduler with the state-of-the-art LLM serving system and show significant performance improvement in several important applications: 2.8x lower latency in chatbot serving and 6.5x higher throughput in synthetic data generation. Our code is available at https://github.com/hao-ai-lab/vllm-ltr.git

Large language models (LLMs) have shown impressive performance on language tasks but face challenges when deployed on resource-constrained devices due to their extensive parameters and reliance on dense multiplications, resulting in high memory demands and latency bottlenecks. Shift-and-add reparameterization offers a promising solution by replacing costly multiplications with hardware-friendly primitives in both the attention and multi-layer perceptron (MLP) layers of an LLM. However, current reparameterization techniques require training from scratch or full parameter fine-tuning to restore accuracy, which is resource-intensive for LLMs. To address this, we propose accelerating pretrained LLMs through post-training shift-and-add reparameterization, creating efficient multiplication-free models, dubbed ShiftAddLLM. Specifically, we quantize each weight matrix into binary matrices paired with group-wise scaling factors. The associated multiplications are reparameterized into (1) shifts between activations and scaling factors and (2) queries and adds according to the binary matrices. To reduce accuracy loss, we present a multi-objective optimization method to minimize both weight and output activation reparameterization errors. Additionally, based on varying sensitivity across layers to reparameterization, we develop an automated bit allocation strategy to further reduce memory usage and latency. Experiments on five LLM families and eight tasks consistently validate the effectiveness of ShiftAddLLM, achieving average perplexity reductions of 5.6 and 22.7 points at comparable or lower latency compared to the most competitive quantized LLMs at 3- and 2-bit precision, respectively, and more than 80% memory and energy reductions over the original LLMs. Codes and models are available at https://github.com/GATECH-EIC/ShiftAddLLM.

Deep neural networks (DNNs) have been widely deployed in software to address various tasks (e.g., autonomous driving, medical diagnosis). However, they can also produce incorrect behaviors that result in financial losses and even threaten human safety. To reveal and repair incorrect behaviors in DNNs, developers often collect rich, unlabeled datasets from the natural world and label them to test DNN models. However, properly labeling a large number of datasets is a highly expensive and time-consuming task. To address the above-mentioned problem, we propose neuron sensitivity-guided test case selection (NSS), which can reduce the labeling time by selecting valuable test cases from unlabeled datasets. NSS leverages the information of the internal neuron induced by the test cases to select valuable test cases, which have high confidence in causing the model to behave incorrectly. We evaluated NSS with four widely used datasets and four well-designed DNN models compared to the state-of-the-art (SOTA) baseline methods. The results show that NSS performs well in assessing the probability of failure triggering in test cases and in the improvement capabilities of the model. Specifically, compared to the baseline approaches, NSS achieves a higher fault detection rate (e.g., when selecting 5% of the test cases from the unlabeled dataset in the MNIST and LeNet1 experiment, NSS can obtain an 81.8% fault detection rate, which is a 20% increase compared with SOTA baseline strategies).