Ziang Chen, Guannan Li, Fanlin Ji, Yipeng Kang, Jiaqi Li, Muhan Zhang, Yangtao Zhang, Li Tianjiao, Jiannan Wang, Xin Guo, Song-Chun Zhu, Bin Ling. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

In scenarios where language models must incorporate new information efficiently without extensive retraining, traditional fine-tuning methods are prone to overfitting, degraded generalization, and unnatural language generation. To address these limitations, we introduce Consistent In-Context Editing (ICE), a novel approach leveraging the model's in-context learning capability to optimize towards a contextual distribution rather than a one-hot target. ICE introduces a simple yet effective optimization framework for the model to internalize new knowledge by aligning its output distributions with and without additional context. This method enhances the robustness and effectiveness of gradient-based tuning methods, preventing overfitting and preserving the model's integrity. We analyze ICE across four critical aspects of knowledge editing: accuracy, locality, generalization, and linguistic quality, demonstrating its advantages. Experimental results confirm the effectiveness of ICE and demonstrate its potential for continual editing, ensuring that the integrity of the model is preserved while updating information.

Large language models (LLMs) are typically limited to processing texts within contextwindow size, which has spurred significant research efforts into enhancing LLMs' longcontext understanding as well as developing high-quality benchmarks to evaluate the ability.However, prior datasets suffer from shortcomings like short length compared to the context window of modern LLMs; outdated documents that might have data leakage problems; and an emphasis on short dependency tasks only.In this paper, we present ooGLE , a Long Context Generic Language Evaluation benchmark.It features documents post-2022, with over 24,000 tokens per document and 6,000 newly generated questions spanning varying dependency ranges in diverse domains.Human annotators meticulously crafted over 1,100 high-quality question-answer (QA) pairs with thorough cross-validation for a most precise assessment of LLMs' long dependency capabilities.We conduct a comprehensive evaluation of representative LLMs on ooGLE .The results indicate that most LLMs have shockingly bad long context ability and fail to capture long dependencies in the context, even when their context window size is enough to fit the entire document.Our results shed light on enhancing the "true long-context understanding" ability of LLMs instead of merely enlarging their context window.

Large Language Models (LLMs) trained on massive corpora have shown remarkable success in knowledge-intensive tasks. Yet, most of them rely on pre-stored knowledge. Inducing new general knowledge from a specific environment andperforming reasoning with the acquired knowledge—situated inductive reasoning, is crucial and challenging for machine intelligence. In this paper, we design Mars, an interactive environment devised for situated inductive reasoning. It introduces counter-commonsense game mechanisms by modifying terrain, survival setting and task dependency while adhering to certain principles. In Mars, agents need to actively interact with their surroundings, derive useful rules and perform decision-making tasks in specific contexts. We conduct experiments on various RL-based and LLM-based methods, finding that they all struggle on this challenging situated inductive reasoning benchmark. Furthermore, we explore Induction from Reflection, where we instruct agents to perform inductive reasoning from history trajectory. The superior performance underscores the importance of inductive reasoning in Mars. Through Mars, we aim to galvanize advancements in situated inductive reasoning and set the stage for developing the next generation of AI systems that can reason in an adaptive and context-sensitive way.

The availability of online services is vital as its strong relevance to revenue and user experience. To ensure online services’ availability, quickly localizing the root causes of system failures is crucial. Given the high resource consumption of traces, call metric data are widely used by existing approaches to construct call graphs in practice. However, ambiguous correspondences between upstream and downstream calls may exist and result in exploring unexpected edges in the constructed call graph. Conducting root cause localization on this graph may lead to misjudgments of real root causes. To the best of our knowledge, we are the first to investigate such ambiguity, which is overlooked in the existing literature. Inspired by the law of large numbers and the Markov properties of network traffic, we propose a regression-based method (named AmSitor) to address this problem effectively. Based on AmSitor, we propose an ambiguity-aware root cause localization approach based on Call Metric Data named CMDiagnostor, containing metric anomaly detection, ambiguity-free call graph construction, root cause exploration, and candidate root cause ranking modules. The comprehensive experimental evaluations conducted on real-world datasets show that our CMDiagnostor can outperform the state-of-the-art approaches by 14% on the top-5 hit rate. Moreover, AmSitor can also be applied to existing baseline approaches separately to improve their performances one step further. The source code is released at https://github.com/NetManAIOps/CMDiagnostor.