The rapid evolution of reasoning-intensive Large Language Models renders traditional metrics insufficient by masking fine-grained failures and implicit pathologies. Existing weakness discovery methods typically rely on rigid pipelines, yielding superficial insights that lack the diagnostic depth required for effective model improvement. To address this, we introduce Agent4Weakness, a multi-agent framework designed to replicate the rigorous workflow of human expert analysts. By integrating a Domain-Aware Memory for contextual reasoning with diagnostic evaluation skills and a Tool Abstraction mechanism for decouple data analysis, Agent4Weakness transforms raw evaluation traces into grounded, actionable reports. We validate our framework through an extensive study involving 104 models across 27 benchmarks. Experimental results demonstrate that Agent4Weakness produces diagnostic reports significantly superior to competitive baselines. Crucially, leveraging these insights for prompt guidance yields an average 3.7 point performance boost and establishes a closed-loop optimization framework.

Large Language Models (LLMs) perform well on standard reasoning and question-answering benchmarks, yet such evaluations often fail to capture their ability to handle long-tail, expertise-intensive knowledge in real-world professional scenarios. We introduce LPFQA, a long-tail knowledge benchmark derived from authentic professional forum discussions, covering 7 academic and industrial domains with 430 curated tasks grounded in practical expertise. LPFQA evaluates specialized reasoning, domain-specific terminology understanding, and contextual interpretation, and adopts a hierarchical difficulty structure to ensure semantic clarity and uniquely identifiable answers. Experiments on over multiple mainstream LLMs reveal substantial performance gaps, particularly on tasks requiring deep domain reasoning, exposing limitations overlooked by existing benchmarks. Overall, LPFQA provides an authentic and discriminative evaluation framework that complements prior benchmarks and informs future LLM development. Our data is available at https://huggingface.co/datasets/m-a-p/LPFQA.

Moral judgments in LLMs increasingly capture the attention of researchers in AI ethics domain. This study explores moral judgments of three open-source large language models (LLMs)—Qwen-1.5-14B, Llama3-8B, and DeepSeek-R1 in plausible moral dilemmas, examining their sensitivity to social exposure and collaborative decision-making. Using a dual-process framework grounded in deontology and utilitarianism, we evaluate LLMs’ responses to moral dilemmas under varying social contexts. Results reveal that all models are significantly influenced by moral norms rather than consequences, with DeepSeek-R1 exhibiting a stronger action tendency compared to Qwen-1.5-14B and Llama3-8B, which show higher inaction preferences. Social exposure and collaboration impact LLMs differently: Qwen-1.5-14B becomes less aligned with moral norms under observation, while DeepSeek-R1’s action tendency is moderated by social collaboration. These findings highlight the nuanced moral reasoning capabilities of LLMs and their varying sensitivity to social cues, providing insights into the ethical alignment of AI systems in socially embedded contexts.

Spatial representations are fundamental to human cognition, as understanding spatial relationships between objects is essential in daily life. Language serves as an indispensable tool for communicating spatial information, creating a close connection between spatial representations and spatial language. Large language models (LLMs), theoretically, possess spatial cognition due to their proficiency in natural language processing. This study examines the spatial representations of LLMs by employing traditional spatial tasks used in human experiments and comparing the models’ performance to that of humans. The results indicate that LLMs resemble humans in selecting spatial prepositions to describe spatial relationships and exhibit a preference for vertically oriented spatial terms. However, the human tendency to better represent locations along specific axes is absent in the performance of LLMs. This finding suggests that, although spatial language is closely linked to spatial representations, the two are not entirely equivalent.

Sequential recommendation systems play a key role in modern information retrieval. However, existing intent-related work fails to adequately capture long-term dependencies in user behavior, i.e., the influence of early user behavior on current behavior, and also fails to effectively utilize item relevance. To this end, we propose a novel sequential recommendation framework to overcome the above limitations, called ICMA. Specifically, we combine temporal variability with position encoding that has extrapolation properties to encode sequences, thereby expanding the model’s view of user behavior and capturing long-term user dependencies more effectively. Additionally, we design a multi-view data augmentation method, i.e., based on random data augmentation methods (e.g., crop, mask, and reorder), and further introduce insertion and substitution operations to augment the sequence data from different views by utilizing item relevance. Within this framework, clustering is performed to learn intent distributions, and these learned intents are integrated into the sequential recommendation model via contrastive SSL, which maximizes consistency between sequence views and their corresponding intents. The training process alternates between the Expectation (E) step and the Maximization (M) step. Experiments on three real datasets show that our approach improves by 0.8% to 14.7% compared to most baselines.