Recent advancements have significantly expanded the theoretical context windows of Multimodal Embedding Models (MEMs). However, larger context windows do not necessarily translate into effective comprehension and representation of long-context multimodal inputs, which remains a critical bottleneck for real-world deployment. To address the lack of systematic evaluation in this setting, we introduce *MMLongEmbed*, the first comprehensive benchmark for evaluating MEMs in long-context scenarios. MMLongEmbed comprises four retrieval tasks spanning multiple context-length ranges, covering text, document, and video modalities. Through extensive evaluation of state-of-the-art models, we find that current architectures rely heavily on superficial feature matching and struggle to capture deep semantic and structural dependencies. We further observe that performance degradation varies systematically with context length and key information placement. Moreover, models exhibit substantially different robustness to redundant contextual information across modalities.

Large language models (LLMs) are increasingly used in high-stakes domains, yet their unfaithful or non-factual outputs create serious risks and hinder trustworthy deployment. Rigorous fact-checking is therefore essential to mitigate these risks. Most existing factuality verification methods improve evidence acquisition, but fail to capture complex logical relations between claims and multi-source evidence, especially when different evidence sources are inconsistent or incomplete. Moreover, current benchmarks rarely capture such realistic conflict scenarios. To address this, we introduce MAVEN, a real-world benchmark for verifying complex claims against heterogeneous and conflicting web evidence with fine-grained explanations. We further propose ArgCheck, an argumentation-based fact-checking framework. It models claims and retrieved evidence as arguments linked by attack relations, resolving conflicts through formal argumentation semantics. To avoid incomplete evidence, ArgCheck incorporates adversarial web retrieval to gather balanced multi-source evidence and grounds its verdicts and explanations in accepted arguments. Experiments on MAVEN and the public fact-checking benchmark AVeriTeC show that ArgCheck improves Macro-F1 over the strongest baseline by 5.39 and 11.10 points, respectively, while generating more faithful and evidence-grounded explanations. Our code and dataset are available online https://github.com/ZyraChen/ArgCheck.

Attributed text generation aims to enhance the reliability of content generated from large language models by providing citations for each claim, which thereby enables users to easily verify the correctness of the responses. However, the scarcity of high-quality training samples presents a significant challenge in aligning large language models to generate texts with citations, revealing considerable room for improvement in existing attribution systems. Besides, existing approaches of aligning large language models to follow user instructions can lead to an undue emphasis on irrelevant documents, which in turn reduces the quality of responses. To address the above problems, we propose Automatic Alignment Framework for Attributed Text Generation (𝒜³), a novel framework designed to automatically generate high-quality attributed query-response pairs for both supervised fine-tuning and preference optimization stages without human annotation. With the help of 𝒜³, Mistral-7B can achieve a citation recall of 84.4 and a precision of 87.0 precision on ASQA, which notably surpasses GPT-4’s citation recall of 73.0 and precision of 76.5.

Large Language Models (LLMs) hold considerable promise for artificial general intelligence, given their intrinsic abilities to accomplish a wide range of open-domain tasks either independently or in tandem with specialized expert models. However, despite these capabilities, the performance of LLMs has yet to be comprehensively evaluated in realistic scenarios. To this end, in this work, we introduce a novel task, the Realistic Chinese Spell Checking (RCSC), to evaluate the effectiveness of existing methods comprehensively. In contrast to existing works that solely address Chinese character misspellings or pinyin conversions, our task aims to convert the realistic Chinese text into the corresponding correct text. The realistic Chinese text may potentially contain both Chinese misspellings and pinyin conversions. We first present the Realistic Chinese Spell Checking Benchmark (RCSCB), which consists of two subsets and contains a total of 581,657 samples. Then, we benchmark the performance of various baselines and find that all the existing methods, including instruction-based LLMs, achieve unsatisfactory results on RCSCB. To further improve the performance on RCSCB, we propose Pinyin-Enhanced Spell Checker (PESC), which is specifically designed to address pinyin-related misspellings. Experimental results demonstrate that PESC can achieve state-of-the-art performance on RCSCB. Despite the progress made, the current state-of-the-art performance is still far from satisfactory. We expect further progress on this crucial and challenging task.

Hierarchical text classification (HTC) focuses on classifying one text into multiple labels, which are organized as a hierarchical taxonomy. Due to its wide involution in realistic scenarios, HTC attracts long-term attention from both industry and academia. However, the high cost of hierarchical multi-label annotation makes HTC suffer from the data scarcity problem. In view of the difficulty in balancing the controllability of multiple structural labels and text diversity, automatically generating high-quality data for HTC is challenging and under-explored. To fill this blank, we propose a novel data generation framework tailored for HTC, which can achieve both label controllability and text diversity by extracting high-quality semantic-level and phrase-level hierarchical label information. Experimental results on three benchmarks demonstrate that, compared with existing data augmentation methods, the data generated from our method can bring the most significant performance improvements of several strong HTC models. Extensive analysis confirms that the improvements yielded by our proposed method do correlate to the enhancement of label controllability and text diversity.

Contrastive learning has achieved impressive success in generation tasks to militate the “exposure bias” problem and discriminatively exploit the different quality of references. Existing works mostly focus on contrastive learning on the instance-level without discriminating the contribution of each word, while keywords are the gist of the text and dominant the constrained mapping relationships. Hence, in this work, we propose a hierarchical contrastive learning mechanism, which can unify hybrid granularities semantic meaning in the input text. Concretely, we first propose a keyword graph via contrastive correlations of positive-negative pairs to iteratively polish the keyword representations. Then, we construct intra-contrasts within instance-level and keyword-level, where we assume words are sampled nodes from a sentence distribution. Finally, to bridge the gap between independent contrast levels and tackle the common contrast vanishing problem, we propose an inter-contrast mechanism that measures the discrepancy between contrastive keyword nodes respectively to the instance distribution. Experiments demonstrate that our model outperforms competitive baselines on paraphrasing, dialogue generation, and storytelling tasks.