Understanding LLMs' Cross-Lingual Context Retrieval: How Good It Is And Where It Comes From

The ability of cross-lingual context retrieval is a fundamental aspect of cross-lingual alignment of large language models (LLMs), where the model extracts context information in one language based on requests in another language. Despite its importance in real-life applications, this ability has not been adequately investigated for state-of-the-art models. In this paper, we evaluate the cross-lingual context retrieval ability of over 40 LLMs across 12 languages to understand the source of this ability, using cross-lingual machine reading comprehension (xMRC) as a representative scenario. Our results show that several small, post-trained open LLMs show strong cross-lingual context retrieval ability, comparable to closed-source LLMs such as GPT-4o, and their estimated oracle performances greatly improve after post-training. Our interpretability analysis shows that the cross-lingual context retrieval process can be divided into two main phases: question encoding and answer retrieval, which are formed in pre-training and post-training, respectively. The phasing stability correlates with xMRC performance, and the xMRC bottleneck lies at the last model layers in the second phase, where the effect of post-training can be evidently observed. Our results also indicate that larger-scale pretraining cannot improve the xMRC performance. Instead, larger LLMs need further multilingual post-training to fully unlock their cross-lingual context retrieval potential. Our code and is available at https://github.com/NJUNLP/Cross-Lingual-Context-Retrieval

A Cross-Domain Few-Shot Learning Method Based on Domain Knowledge Mapping

In task-based few-shot learning paradigms, it is commonly assumed that different tasks are independently and identically distributed (i.i.d.). However, in real-world scenarios, the distribution encountered in few-shot learning can significantly differ from the distribution of existing data. Thus, how to effectively leverage existing data knowledge to enable models to quickly adapt to class variations under non-i.i.d. assumptions has emerged as a key research challenge. To address this challenge, this paper proposes a new cross-domain few-shot learning approach based on domain knowledge mapping, applied consistently throughout the pre-training, training, and testing phases. In the pre-training phase, our method integrates self-supervised and supervised losses by maximizing mutual information, thereby mitigating mode collapse. During the training phase, the domain knowledge mapping layer collaborates with a domain classifier to learn both domain mapping capabilities and the ability to assess domain adaptation difficulty. Finally, this approach is applied during the testing phase, rapidly adapting to domain variations through meta-training tasks on support sets, consequently enhancing the model's capability to transfer domain knowledge effectively. Experimental validation conducted across six datasets from diverse domains demonstrates the effectiveness of the proposed method.

FamilyTool: A Multi-hop Personalized Tool Use Benchmark

The integration of tool learning with Large Language Models (LLMs) has expanded their capabilities in handling complex tasks by leveraging external tools. However, existing benchmarks for tool learning inadequately address critical real-world personalized scenarios, particularly those requiring multi-hop reasoning and inductive knowledge adaptation in dynamic environments. To bridge this gap, we introduce FamilyTool, a novel benchmark grounded in a family-based knowledge graph (KG) that simulates personalized, multi-hop tool use scenarios. FamilyTool challenges LLMs with queries spanning 1 to 3 relational hops (e.g., inferring familial connections and preferences) and incorporates an inductive KG setting where models must adapt to unseen user preferences and relationships without re-training, a common limitation in prior approaches that compromises generalization. We further propose KGETool: a simple KG-augmented evaluation pipeline to systematically assess LLMs' tool use ability in these settings. Experiments reveal significant performance gaps in state-of-the-art LLMs, with accuracy dropping sharply as hop complexity increases and inductive scenarios exposing severe generalization deficits. These findings underscore the limitations of current LLMs in handling personalized, evolving real-world contexts and highlight the urgent need for advancements in tool-learning frameworks. FamilyTool serves as a critical resource for evaluating and advancing LLM agents' reasoning, adaptability, and scalability in complex, dynamic environments. Code and dataset are available at Github.

R-PRM: Reasoning-Driven Process Reward Modeling

Large language models (LLMs) inevitably make mistakes when performing step-by-step mathematical reasoning. Process Reward Models (PRMs) have emerged as a promising solution by evaluating each reasoning step. However, existing PRMs typically output evaluation scores directly, limiting both learning efficiency and evaluation accuracy, which is further exacerbated by the scarcity of annotated data. To address these issues, we propose Reasoning-Driven Process Reward Modeling (R-PRM). First, we leverage stronger LLMs to generate seed data from limited annotations, effectively bootstrapping our model's reasoning capabilities and enabling comprehensive step-by-step evaluation. Second, we further enhance performance through preference optimization, without requiring additional annotated data. Third, we introduce inference-time scaling to fully harness the model's reasoning potential. Extensive experiments demonstrate R-PRM's effectiveness: on ProcessBench and PRMBench, it surpasses strong baselines by 11.9 and 8.5 points in F1 scores, respectively. When applied to guide mathematical reasoning, R-PRM achieves consistent accuracy improvements of over 8.5 points across six challenging datasets. Further analysis reveals that R-PRM exhibits more comprehensive evaluation and stronger generalization capabilities, thereby highlighting its significant potential.

CapArena: Benchmarking and Analyzing Detailed Image Captioning in the LLM Era

Image captioning has been a longstanding challenge in vision-language research. With the rise of LLMs, modern Vision-Language Models (VLMs) generate detailed and comprehensive image descriptions. However, benchmarking the quality of such captions remains unresolved. This paper addresses two key questions: (1) How well do current VLMs actually perform on image captioning, particularly compared to humans? We built CapArena, a platform with over 6000 pairwise caption battles and high-quality human preference votes. Our arena-style evaluation marks a milestone, showing that leading models like GPT-4o achieve or even surpass human performance, while most open-source models lag behind. (2) Can automated metrics reliably assess detailed caption quality? Using human annotations from CapArena, we evaluate traditional and recent captioning metrics, as well as VLM-as-a-Judge. Our analysis reveals that while some metrics (e.g., METEOR) show decent caption-level agreement with humans, their systematic biases lead to inconsistencies in model ranking. In contrast, VLM-as-a-Judge demonstrates robust discernment at both the caption and model levels. Building on these insights, we release CapArena-Auto, an accurate and efficient automated benchmark for detailed captioning, achieving 94.3% correlation with human rankings at just $4 per test. Data and resources will be open-sourced at https://caparena.github.io.

Alleviating Distribution Shift in Synthetic Data for Machine Translation Quality Estimation

Quality Estimation (QE) models evaluate the quality of machine translations without reference translations, serving as the reward models for the translation task. Due to the data scarcity, synthetic data generation has emerged as a promising solution. However, synthetic QE data often suffers from distribution shift, which can manifest as discrepancies between pseudo and real translations, or in pseudo labels that do not align with human preferences. To tackle this issue, we introduce ADSQE, a novel framework for alleviating distribution shift in synthetic QE data. To reduce the difference between pseudo and real translations, we employ the constrained beam search algorithm and enhance translation diversity through the use of distinct generation models. ADSQE uses references, i.e., translation supervision signals, to guide both the generation and annotation processes, enhancing the quality of word-level labels. ADSE further identifies the shortest phrase covering consecutive error tokens, mimicking human annotation behavior, to assign the final phrase-level labels. Specially, we underscore that the translation model can not annotate translations of itself accurately. Extensive experiments demonstrate that ADSQE outperforms SOTA baselines like COMET in both supervised and unsupervised settings. Further analysis offers insights into synthetic data generation that could benefit reward models for other tasks.

Generalizing From Short to Long: Effective Data Synthesis for Long-Context Instruction Tuning

Long-context modelling for large language models (LLMs) has been a key area of recent research because many real world use cases require reasoning over longer inputs such as documents. The focus of research into modelling long context has been on how to model position and there has been little investigation into other important aspects of language modelling such as instruction tuning. Long context training examples are challenging and expensive to create and use. In this paper, we investigate how to design instruction data for the post-training phase of a long context pre-trained model: how much and what type of context is needed for optimal and efficient post-training. Our controlled study reveals that models instruction-tuned on short contexts can effectively generalize to longer ones, while also identifying other critical factors such as instruction difficulty and context composition. Based on these findings, we propose context synthesis, a novel data synthesis framework that leverages off-the-shelf LLMs to generate extended background contexts for high-quality instruction-answer pairs. Experiment results on the document-level benchmark (LongBench) demonstrate that our proposed approach outperforms previous instruction synthesis approaches and comes close to the performance of human-annotated long-context instruction data. The project will be available at: https://github.com/NJUNLP/context-synthesis.

Extend Adversarial Policy Against Neural Machine Translation via Unknown Token

Generating adversarial examples contributes to mainstream neural machine translation~(NMT) robustness. However, popular adversarial policies are apt for fixed tokenization, hindering its efficacy for common character perturbations involving versatile tokenization. Based on existing adversarial generation via reinforcement learning~(RL), we propose the `DexChar policy' that introduces character perturbations for the existing mainstream adversarial policy based on token substitution. Furthermore, we improve the self-supervised matching that provides feedback in RL to cater to the semantic constraints required during training adversaries. Experiments show that our method is compatible with the scenario where baseline adversaries fail, and can generate high-efficiency adversarial examples for analysis and optimization of the system.

Rethinking Relation Extraction: Beyond Shortcuts to Generalization with a Debiased Benchmark

Benchmarks are crucial for evaluating machine learning algorithm performance, facilitating comparison and identifying superior solutions. However, biases within datasets can lead models to learn shortcut patterns, resulting in inaccurate assessments and hindering real-world applicability. This paper addresses the issue of entity bias in relation extraction tasks, where models tend to rely on entity mentions rather than context. We propose a debiased relation extraction benchmark DREB that breaks the pseudo-correlation between entity mentions and relation types through entity replacement. DREB utilizes Bias Evaluator and PPL Evaluator to ensure low bias and high naturalness, providing a reliable and accurate assessment of model generalization in entity bias scenarios. To establish a new baseline on DREB, we introduce MixDebias, a debiasing method combining data-level and model training-level techniques. MixDebias effectively improves model performance on DREB while maintaining performance on the original dataset. Extensive experiments demonstrate the effectiveness and robustness of MixDebias compared to existing methods, highlighting its potential for improving the generalization ability of relation extraction models. We will release DREB and MixDebias publicly.

Enhancing Mathematical Reasoning in LLMs with Background Operators

We propose utilizing background operators for mathematical reasoning in large language models (LLMs). To achieve this, we define a set of fundamental mathematical predicates as the basic building blocks. For each mathematical problem, we develop a Prolog solution that includes problem-specific predicates and intermediate predicates derived from these background operators, ensuring that each solution adheres to the defined operator set. We introduce the MATH-Prolog corpus, which is derived from the counting and probability categories of the MATH corpus. For efficient data augmentation, we apply K-fold cross-validated self-training. This method incrementally generates new Prolog solutions for each fold, incorporating those verified as correct into the training set throughout the model training process. Our experimental results demonstrate that 5-fold crossvalidated self-training effectively identifies new, accurate Prolog solutions, achieving an accuracy of 84.6% on the cross-validated set, and 84.8% on the test set during fine-tuning the Meta-Llama-3.1-8B-Instruct model. This approach successfully uncovers new solutions with fully computable inference steps for previously unseen problems. Additionally, incorporating the background mathematical predicates into the prompt enhances solution coverage.