Jianzhu Bao, Haozhen Zhang, Kuicai Dong, Bozhi Wu, Sarthak Ketanbhai Modi, Zi Pong Lim, Yon Shin Teo, Wenya Wang. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Abstract Requirement defects are a major source of late-stage failures in automotive systems, yet rigorous validation is rarely applied during early development. While formal methods offer strong guarantees, their adoption at the requirements level is limited by high formalization cost and expertise barriers. We present an industry-oriented, shift-left verification approach that integrates Large Language Models (LLMs) with formal methods to enable requirements-level validation. Requirements are classified and decomposed by LLMs, translated into CSP system models and assertions, and refined through a CEGAR-inspired loop using the FDR4 model checker. Validation is decomposed into requirement–assertion pairs supported by natural-language back-translations and confidence scores, preserving expert control without manual formal modeling. Domain knowledge–based validation further leverages historical defect data to identify implicit requirement gaps. We evaluate the approach on three real-world automotive case studies. Results show high automation for small-to-medium systems (80–100% synthesis success for up to $$\sim 60$$ ∼ 60 requirements), effective expert validation guided by LLM confidence estimates, and 100% detection of known historical defects alongside 22 novel gaps. The workflow completes within 10–35 min per project at negligible cost ( $$<8$$ < 8 per project), with limited expert effort. Our results demonstrate that LLM+formal method hybridization can provide scalable, rigorous, and industrially viable requirements-level verification, supporting practical shift-left adoption in automotive systems.

Deep reinforcement learning (DRL) has revolutionized quantitative trading (Q-trading) by achieving decent performance without significant human expert knowledge. Despite its achievements, we observe that the current state-of-the-art DRL models are still ineffective in identifying the market trends, causing them to miss good trading opportunity or suffer from large drawdowns when encountering market crashes. To address this limitation, a natural approach is to incorporate human expert knowledge in identifying market trends. Whereas, such knowledge is abstract and hard to be quantified. In order to effectively leverage abstract human expert knowledge, in this paper, we propose a universal logic-guided deep reinforcement learning framework for Q-trading, called Logic-Q. In particular, Logic-Q adopts the program synthesis by sketching paradigm and introduces a logic-guided model design that leverages a lightweight, plug-and-play market trend-aware program sketch to determine the market trend and correspondingly adjusts the DRL policy in a post-hoc manner. Extensive evaluations of two popular quantitative trading tasks demonstrate that Logic-Q can significantly improve the performance of previous state-of-the-art DRL trading strategies.

Deep learning has introduced significant improvements in many software analysis tasks. Although the Large Language Models (LLMs) based neural code models demonstrate commendable performance when trained and tested within the intra-project independent and identically distributed (IID) setting, they often struggle to generalize effectively to real-world inter-project out-of-distribution (OOD) data. In this work, we show that this phenomenon is caused by the heavy reliance on project-specific shortcuts for prediction instead of ground-truth evidence. We propose a Cond-Idf measurement to interpret this behavior, which quantifies the relatedness of a token with a label and its project-specificness. The strong correlation between model behavior and the proposed measurement indicates that without proper regularization, models tend to leverage spurious statistical cues for prediction. Equipped with these observations, we propose a novel bias mitigation mechanism that regularizes the model’s learning behavior by leveraging latent logic relations among samples. Experimental results on two representative program analysis tasks indicate that our mitigation framework can improve both inter-project OOD generalization and adversarial robustness, while not sacrificing accuracy on intra-project IID data.

Reasoning is a fundamental ability towards artificial general intelligence (AGI). Fueled by the success of deep learning, the neural logic machines models (NLMs) have introduced novel neural-symbolic structures and demonstrate great performance and generalization on reasoning and decision-making tasks. However, the original training approaches of the NLMs are still far from perfect, the models would repeat similar mistakes during the training process which leads to sub-optimal performance. To mitigate this issue, we present a novel framework named Failure Reflection Guided Regularizer (FRGR). FRGR first dynamically identifies and summarizes the root cause if the model repeats similar mistakes during training. Then it penalizes the model if it makes similar mistakes in future training iterations. In this way, the model is expected to avoid repeating errors of similar root causes and converge faster to a better-performed optimum. Experimental results on multiple relational reasoning and decision-making tasks demonstrate the effectiveness of FRGR in improving performance, generalization, training efficiency, and data efficiency.

Large-scale pre-trained models such as CodeBERT, GraphCodeBERT have earned widespread attention from both academia and industry. Attributed to the superior ability in code representation, they have been further applied in multiple downstream tasks such as clone detection, code search and code translation. However, it is also observed that these state-of-the-art pre-trained models are susceptible to adversarial attacks. The performance of these pre-trained models drops significantly with simple perturbations such as renaming variable names. This weakness may be inherited by their downstream models and thereby amplified at an unprecedented scale. To this end, we propose an approach namely ContraBERT that aims to improve the robustness of pre-trained models via contrastive learning. Specifically, we design nine kinds of simple and complex data augmentation operators on the programming language (PL) and natural language (NL) data to construct different variants. Furthermore, we continue to train the existing pre-trained models by masked language modeling (MLM) and contrastive pre-training task on the original samples with their augmented variants to enhance the robustness of the model. The extensive ex-periments demonstrate that ContraBERT can effectively improve the robustness of the existing pre-trained models. Further study also confirms that these robustness-enhanced models provide improvements as compared to original models over four popular downstream tasks.

Learning-based approaches that learn code representations for software vulnerability detection have been proven to produce inspiring results. However, they still fail to capture complete and precise vulnerability semantics for code representations. To address the limitations, in this work, we propose a learning-based approach namely SnapVuln, which first utilizes multiple vulnerability-specific inter-procedural slicing algorithms to capture vulnerability semantics of various types and then employs a Gated Graph Neural Network (GGNN) with an attention mechanism to learn vulnerability semantics. We compare SnapVuln with state-of-the-art learning-based approaches on two public datasets, and confirm that SnapVuln outperforms them. We further perform an ablation study and demonstrate that the completeness and precision of vulnerability semantics captured by SnapVuln contribute to the performance improvement.

Machine learning–(ML) based approach is considered as one of the most promising techniques for Android malware detection and has achieved high accuracy by leveraging commonly used features. In practice, most of the ML classifications only provide a binary label to mobile users and app security analysts. However, stakeholders are more interested in the reason why apps are classified as malicious in both academia and industry. This belongs to the research area of interpretable ML but in a specific research domain (i.e., mobile malware detection). Although several interpretable ML methods have been exhibited to explain the final classification results in many cutting-edge Artificial Intelligent–based research fields, until now, there is no study interpreting why an app is classified as malware or unveiling the domain-specific challenges. In this article, to fill this gap, we propose a novel and interpretable ML-based approach (named XMal ) to classify malware with high accuracy and explain the classification result meanwhile. (1) The first classification phase of XMal hinges multi-layer perceptron and attention mechanism and also pinpoints the key features most related to the classification result. (2) The second interpreting phase aims at automatically producing neural language descriptions to interpret the core malicious behaviors within apps. We evaluate the behavior description results by leveraging a human study and an in-depth quantitative analysis. Moreover, we further compare XMal with the existing interpretable ML-based methods (i.e., Drebin and LIME) to demonstrate the effectiveness of XMal . We find that XMal is able to reveal the malicious behaviors more accurately. Additionally, our experiments show that XMal can also interpret the reason why some samples are misclassified by ML classifiers. Our study peeks into the interpretable ML through the research of Android malware detection and analysis.