Rethinking Domain-Specific LLM Benchmark Construction: A Comprehensiveness-Compactness Approach

Numerous benchmarks have been built to evaluate the domain-specific abilities of large language models (LLMs), highlighting the need for effective and efficient benchmark construction. Existing domain-specific benchmarks primarily focus on the scaling law, relying on massive corpora for supervised fine-tuning or generating extensive question sets for broad coverage. However, the impact of corpus and question-answer (QA) set design on the precision and recall of domain-specific LLMs remains unexplored. In this paper, we address this gap and demonstrate that the scaling law is not always the optimal principle for benchmark construction in specific domains. Instead, we propose Comp-Comp, an iterative benchmarking framework based on a comprehensiveness-compactness principle. Here, comprehensiveness ensures semantic recall of the domain, while compactness enhances precision, guiding both corpus and QA set construction. To validate our framework, we conducted a case study in a well-renowned university, resulting in the creation of XUBench, a large-scale and comprehensive closed-domain benchmark. Although we use the academic domain as the case in this work, our Comp-Comp framework is designed to be extensible beyond academia, providing valuable insights for benchmark construction across various domains.

Advancing Dialectal Arabic to Modern Standard Arabic Machine Translation

Dialectal Arabic (DA) poses a persistent challenge for natural language processing (NLP), as most everyday communication in the Arab world occurs in dialects that diverge significantly from Modern Standard Arabic (MSA). This linguistic divide limits access to digital services and educational resources and impedes progress in Arabic machine translation. This paper presents two core contributions to advancing DA-MSA translation for the Levantine, Egyptian, and Gulf dialects, particularly in low-resource and computationally constrained settings: a comprehensive evaluation of training-free prompting techniques, and the development of a resource-efficient fine-tuning pipeline. Our evaluation of prompting strategies across six large language models (LLMs) found that few-shot prompting consistently outperformed zero-shot, chain-of-thought, and our proposed Ara-TEaR method. GPT-4o achieved the highest performance across all prompting settings. For fine-tuning, a quantized Gemma2-9B model achieved a CHrF++ score of 49.88, outperforming zero-shot GPT-4o (44.58). Joint multi-dialect trained models outperformed single-dialect counterparts by over 10% CHrF++, and 4-bit quantization reduced memory usage by 60% with less than 1% performance loss. The results and insights of our experiments offer a practical blueprint for improving dialectal inclusion in Arabic NLP, showing that high-quality DA-MSA machine translation is achievable even with limited resources and paving the way for more inclusive language technologies.

Comparing Apples to Oranges: A Dataset & Analysis of LLM Humour Understanding from Traditional Puns to Topical Jokes

Humour, as a complex language form, is derived from myriad aspects of life, whilst existing work on computational humour has focussed almost exclusively on short pun-based jokes. In this work, we investigate whether the ability of Large Language Models (LLMs) to explain humour depends on the particular humour form. We compare models on simple puns and more complex topical humour that requires knowledge of real-world entities and events. In doing so, we curate a dataset of 600 jokes split across 4 joke types and manually write high-quality explanations. These jokes include heterographic and homographic puns, contemporary internet humour, and topical jokes, where understanding relies on reasoning beyond "common sense", rooted instead in world knowledge regarding news events and pop culture. Using this dataset, we compare the zero-shot abilities of a range of LLMs to accurately and comprehensively explain jokes of different types, identifying key research gaps in the task of humour explanation. We find that none of the tested models (inc. reasoning models) are capable of reliably generating adequate explanations of all joke types, further highlighting the narrow focus of most works in computational humour on overly simple joke forms.

First Return, Entropy-Eliciting Explore

Reinforcement Learning from Verifiable Rewards (RLVR) improves the reasoning abilities of Large Language Models (LLMs) but it struggles with unstable exploration. We propose FR3E (First Return, Entropy-Eliciting Explore), a structured exploration framework that identifies high-uncertainty decision points in reasoning trajectories and performs targeted rollouts to construct semantically grounded intermediate feedback. Our method provides targeted guidance without relying on dense supervision. Empirical results on mathematical reasoning benchmarks(AIME24) show that FR3E promotes more stable training, produces longer and more coherent responses, and increases the proportion of fully correct trajectories. These results highlight the framework's effectiveness in improving LLM reasoning through more robust and structured exploration.

Exploring Task Performance with Interpretable Models via Sparse Auto-Encoders

Large Language Models (LLMs) are traditionally viewed as black-box algorithms, therefore reducing trustworthiness and obscuring potential approaches to increasing performance on downstream tasks. In this work, we apply an effective LLM decomposition method using a dictionary-learning approach with sparse autoencoders. This helps extract monosemantic features from polysemantic LLM neurons. Remarkably, our work identifies model-internal misunderstanding, allowing the automatic reformulation of the prompts with additional annotations to improve the interpretation by LLMs. Moreover, this approach demonstrates a significant performance improvement in downstream tasks, such as mathematical reasoning and metaphor detection.

A Survey on Latent Reasoning

Large Language Models (LLMs) have demonstrated impressive reasoning capabilities, especially when guided by explicit chain-of-thought (CoT) reasoning that verbalizes intermediate steps. While CoT improves both interpretability and accuracy, its dependence on natural language reasoning limits the model's expressive bandwidth. Latent reasoning tackles this bottleneck by performing multi-step inference entirely in the model's continuous hidden state, eliminating token-level supervision. To advance latent reasoning research, this survey provides a comprehensive overview of the emerging field of latent reasoning. We begin by examining the foundational role of neural network layers as the computational substrate for reasoning, highlighting how hierarchical representations support complex transformations. Next, we explore diverse latent reasoning methodologies, including activation-based recurrence, hidden state propagation, and fine-tuning strategies that compress or internalize explicit reasoning traces. Finally, we discuss advanced paradigms such as infinite-depth latent reasoning via masked diffusion models, which enable globally consistent and reversible reasoning processes. By unifying these perspectives, we aim to clarify the conceptual landscape of latent reasoning and chart future directions for research at the frontier of LLM cognition. An associated GitHub repository collecting the latest papers and repos is available at: https://github.com/multimodal-art-projection/LatentCoT-Horizon/.

EvolvTrip: Enhancing Literary Character Understanding with Temporal Theory-of-Mind Graphs

A compelling portrayal of characters is essential to the success of narrative writing. For readers, appreciating a character's traits requires the ability to infer their evolving beliefs, desires, and intentions over the course of a complex storyline, a cognitive skill known as Theory-of-Mind (ToM). Performing ToM reasoning in prolonged narratives requires readers to integrate historical context with current narrative information, a task at which humans excel but Large Language Models (LLMs) often struggle. To systematically evaluate LLMs' ToM reasoning capability in long narratives, we construct LitCharToM, a benchmark of character-centric questions across four ToM dimensions from classic literature. Further, we introduce EvolvTrip, a perspective-aware temporal knowledge graph that tracks psychological development throughout narratives. Our experiments demonstrate that EvolvTrip consistently enhances performance of LLMs across varying scales, even in challenging extended-context scenarios. EvolvTrip proves to be particularly valuable for smaller models, partially bridging the performance gap with larger LLMs and showing great compatibility with lengthy narratives. Our findings highlight the importance of explicit representation of temporal character mental states in narrative comprehension and offer a foundation for more sophisticated character understanding. Our data and code are publicly available at https://github.com/Bernard-Yang/EvolvTrip.

Scaling Test-time Compute for LLM Agents

Scaling test time compute has shown remarkable success in improving the reasoning abilities of large language models (LLMs). In this work, we conduct the first systematic exploration of applying test-time scaling methods to language agents and investigate the extent to which it improves their effectiveness. Specifically, we explore different test-time scaling strategies, including: (1) parallel sampling algorithms; (2) sequential revision strategies; (3) verifiers and merging methods; (4)strategies for diversifying rollouts.We carefully analyze and ablate the impact of different design strategies on applying test-time scaling on language agents, and have follow findings: 1. Scaling test time compute could improve the performance of agents. 2. Knowing when to reflect is important for agents. 3. Among different verification and result merging approaches, the list-wise method performs best. 4. Increasing diversified rollouts exerts a positive effect on the agent's task performance.

Overview of the NLPCC 2025 Shared Task: Gender Bias Mitigation Challenge

As natural language processing for gender bias becomes a significant interdisciplinary topic, the prevalent data-driven techniques, such as pre-trained language models, suffer from biased corpus. This case becomes more obvious regarding those languages with less fairness-related computational linguistic resources, such as Chinese. To this end, we propose a Chinese cOrpus foR Gender bIas Probing and Mitigation (CORGI-PM), which contains 32.9k sentences with high-quality labels derived by following an annotation scheme specifically developed for gender bias in the Chinese context. It is worth noting that CORGI-PM contains 5.2k gender-biased sentences along with the corresponding bias-eliminated versions rewritten by human annotators. We pose three challenges as a shared task to automate the mitigation of textual gender bias, which requires the models to detect, classify, and mitigate textual gender bias. In the literature, we present the results and analysis for the teams participating this shared task in NLPCC 2025.

Selecting Demonstrations for Many-Shot In-Context Learning via Gradient Matching

In-Context Learning (ICL) empowers Large Language Models (LLMs) for rapid task adaptation without Fine-Tuning (FT), but its reliance on demonstration selection remains a critical challenge. While many-shot ICL shows promising performance through scaled demonstrations, the selection method for many-shot demonstrations remains limited to random selection in existing work. Since the conventional instance-level retrieval is not suitable for many-shot scenarios, we hypothesize that the data requirements for in-context learning and fine-tuning are analogous. To this end, we introduce a novel gradient matching approach that selects demonstrations by aligning fine-tuning gradients between the entire training set of the target task and the selected examples, so as to approach the learning effect on the entire training set within the selected examples. Through gradient matching on relatively small models, e.g., Qwen2.5-3B or Llama3-8B, our method consistently outperforms random selection on larger LLMs from 4-shot to 128-shot scenarios across 9 diverse datasets. For instance, it surpasses random selection by 4% on Qwen2.5-72B and Llama3-70B, and by around 2% on 5 closed-source LLMs. This work unlocks more reliable and effective many-shot ICL, paving the way for its broader application.