Prior work has shown that two NLP evaluation experiments that report results for the same quality criterion name (e.g. Fluency) do not necessarily evaluate the same aspect of quality, and the comparability implied by the name can be misleading. Not knowing when two evaluations are comparable in this sense means we currently lack the ability to draw reliable conclusions about system quality on the basis of multiple, independently conducted evaluations. This in turn hampers the ability of the field to progress scientifically as a whole, a pervasive issue in NLP since its beginning (Sparck Jones, 1981). It is hard to see how the issue of unclear comparability can be fully addressed other than by the creation of a standard set of quality criterion names and definitions that the several hundred quality criterion names actually in use in the field can be mapped to, and grounded in. Taking a strictly descriptive approach, the QCET Quality Criteria for Evaluation Taxonomy derives a standard set of quality criterion names and definitions from three surveys of evaluations reported in NLP, and structures them into a hierarchy where each parent node captures common aspects of its child nodes. We present QCET and the resources it consists of, and discuss its three main uses in (i) establishing comparability of existing evaluations, (ii) guiding the design of new evaluations, and (iii) assessing regulatory compliance.

Large language models (LLMs) have been widely adopted across various applications, leveraging customized system prompts for diverse tasks. Facing potential system prompt leakage risks, model developers have implemented strategies to prevent leakage, primarily by disabling LLMs from repeating their context when encountering known attack patterns. However, it remains vulnerable to new and unforeseen prompt-leaking techniques. In this paper, we first introduce a simple yet effective prompt leaking attack to reveal such risks. Our attack is capable of extracting system prompts from various LLM-based application, even from SOTA LLM models such as GPT-4o or Claude 3.5 Sonnet. Our findings further inspire us to search for a fundamental solution to the problems by having no system prompt in the context. To this end, we propose SysVec, a novel method that encodes system prompts as internal representation vectors rather than raw text. By doing so, SysVec minimizes the risk of unauthorized disclosure while preserving the LLM's core language capabilities. Remarkably, this approach not only enhances security but also improves the model's general instruction-following abilities. Experimental results demonstrate that SysVec effectively mitigates prompt leakage attacks, preserves the LLM's functional integrity, and helps alleviate the forgetting issue in long-context scenarios.

Multi-Turn Long-Form Question Answering (MT-LFQA) is a key application paradigm of Large Language Models (LLMs) in knowledge-intensive domains. However, existing benchmarks are limited to single-turn dialogue, while multi-turn dialogue benchmarks typically assess other orthogonal capabilities rather than knowledge-intensive factuality. To bridge this critical gap, we introduce \textbf{KnowMT-Bench}, the \textit{first-ever} benchmark designed to systematically evaluate MT-LFQA for LLMs across knowledge-intensive fields, including medicine, finance, and law. To faithfully assess the model's real-world performance, KnowMT-Bench employs a dynamic evaluation setting where models generate their own multi-turn dialogue histories given logically progressive question sequences. The factual capability and information delivery efficiency of the \textit{final-turn} answer are then evaluated using a human-validated automated pipeline. Our experiments reveal that multi-turn contexts degrade performance: factual capability declines due to the contextual noise from self-generated histories, while information efficiency drops as models become more verbose with increasing dialogue length. We then investigate mitigation strategies, demonstrating that retrieval-augmented generation (RAG) can effectively alleviate and even reverse this factual degradation. These findings underscore the importance of our benchmark in evaluating and enhancing the conversational factual capabilities of LLMs in real-world knowledge-intensive applications. Code is available at \href{https://github.com/hardenyu21/KnowMT-Bench}{\textcolor{cyan}{\texttt{KnowMT-Bench}}}.

Aligning AI systems with human privacy preferences requires understanding individuals' nuanced disclosure behaviors beyond general norms. Yet eliciting such boundaries remains challenging due to the context-dependent nature of privacy decisions and the complex trade-offs involved. We present an AI-powered elicitation approach that probes individuals' privacy boundaries through a discriminative task. We conducted a between-subjects study that systematically varied communication roles and delegation conditions, resulting in 1,681 boundary specifications from 169 participants for 61 scenarios. We examined how these contextual factors and individual differences influence the boundary specification. Quantitative results show that communication roles influence individuals' acceptance of detailed and identifiable disclosure, AI delegation and individuals' need for privacy heighten sensitivity to disclosed identifiers, and AI delegation results in less consensus across individuals. Our findings highlight the importance of situating privacy preference elicitation within real-world data flows. We advocate using nuanced privacy boundaries as an alignment goal for future AI systems.

Risk categorization in 10-K risk disclosures matters for oversight and investment, yet no public benchmark evaluates unsupervised topic models for this task. We present GRAB, a finance-specific benchmark with 1.61M sentences from 8,247 filings and span-grounded sentence labels produced without manual annotation by combining FinBERT token attention, YAKE keyphrase signals, and taxonomy-aware collocation matching. Labels are anchored in a risk taxonomy mapping 193 terms to 21 fine-grained types nested under five macro classes; the 21 types guide weak supervision, while evaluation is reported at the macro level. GRAB unifies evaluation with fixed dataset splits and robust metrics--Accuracy, Macro-F1, Topic BERTScore, and the entropy-based Effective Number of Topics. The dataset, labels, and code enable reproducible, standardized comparison across classical, embedding-based, neural, and hybrid topic models on financial disclosures.

Language Models (LMs) have significantly advanced natural language processing and enabled remarkable progress across diverse domains, yet their black-box nature raises critical concerns about the interpretability of their internal mechanisms and decision-making processes. This lack of transparency is particularly problematic for adoption in high-stakes domains, where stakeholders need to understand the rationale behind model outputs to ensure accountability. On the other hand, while explainable artificial intelligence (XAI) methods have been well studied for non-LMs, they face many limitations when applied to LMs due to their complex architectures, considerable training corpora, and broad generalization abilities. Although various surveys have examined XAI in the context of LMs, they often fail to capture the distinct challenges arising from the architectural diversity and evolving capabilities of these models. To bridge this gap, this survey presents a comprehensive review of XAI techniques with a particular emphasis on LMs, organizing them according to their underlying transformer architectures: encoder-only, decoder-only, and encoder-decoder, and analyzing how methods are adapted to each while assessing their respective strengths and limitations. Furthermore, we evaluate these techniques through the dual lenses of plausibility and faithfulness, offering a structured perspective on their effectiveness. Finally, we identify open research challenges and outline promising future directions, aiming to guide ongoing efforts toward the development of robust, transparent, and interpretable XAI methods for LMs.

This study investigates the extent to which contemporary Text-to-Image artificial intelligence (AI) models perpetuate gender stereotypes and cultural inaccuracies when generating depictions of professionals in Saudi Arabia. We analyzed 1,006 images produced by ImageFX, DALL-E V3, and Grok for 56 diverse Saudi professions using neutral prompts. Two trained Saudi annotators evaluated each image on five dimensions: perceived gender, clothing and appearance, background and setting, activities and interactions, and age. A third senior researcher adjudicated whenever the two primary raters disagreed, yielding 10,100 individual judgements. The results reveal a strong gender imbalance, with ImageFX outputs being 85\% male, Grok 86.6\% male, and DALL-E V3 96\% male, indicating that DALL-E V3 exhibited the strongest overall gender stereotyping. This imbalance was most evident in leadership and technical roles. Moreover, cultural inaccuracies in clothing, settings, and depicted activities were frequently observed across all three models. Counter-stereotypical images often arise from cultural misinterpretations rather than genuinely progressive portrayals. We conclude that current models mirror societal biases embedded in their training data, generated by humans, offering only a limited reflection of the Saudi labour market's gender dynamics and cultural nuances. These findings underscore the urgent need for more diverse training data, fairer algorithms, and culturally sensitive evaluation frameworks to ensure equitable and authentic visual outputs.

Text-to-image diffusion models (DMs) inadvertently reproduce copyrighted styles and protected visual concepts, raising legal and ethical concerns. Concept erasure has emerged as a safeguard, aiming to selectively suppress such concepts through fine-tuning. However, existing methods do not scale to practical settings where providers must erase multiple and possibly conflicting concepts. The core bottleneck is their reliance on static erasure: a single checkpoint is fine-tuned to remove all target concepts, regardless of the actual erasure needs at inference. This rigid design mismatches real-world usage, where requests vary per generation, leading to degraded erasure success and reduced fidelity for non-target content. We propose DyME, an on-demand erasure framework that trains lightweight, concept-specific LoRA adapters and dynamically composes only those needed at inference. This modular design enables flexible multi-concept erasure, but naive composition causes interference among adapters, especially when many or semantically related concepts are suppressed. To overcome this, we introduce bi-level orthogonality constraints at both the feature and parameter levels, disentangling representation shifts and enforcing orthogonal adapter subspaces. We further develop ErasureBench-H, a new hierarchical benchmark with brand-series-character structure, enabling principled evaluation across semantic granularities and erasure set sizes. Experiments on ErasureBench-H and standard datasets (e.g., CIFAR-100, Imagenette) demonstrate that DyME consistently outperforms state-of-the-art baselines, achieving higher multi-concept erasure fidelity with minimal collateral degradation.

The benefit claims of a product is a critical driver of consumers' purchase behavior. Creating product claims is an intense task that requires substantial time and funding. We have developed the $\textbf{Claim Advisor}$ web application to accelerate claim creations using in-context learning and fine-tuning of large language models (LLM). $\textbf{Claim Advisor}$ was designed to disrupt the speed and economics of claim search, generation, optimization, and simulation. It has three functions: (1) semantically searching and identifying existing claims and/or visuals that resonate with the voice of consumers; (2) generating and/or optimizing claims based on a product description and a consumer profile; and (3) ranking generated and/or manually created claims using simulations via synthetic consumers. Applications in a consumer packaged goods (CPG) company have shown very promising results. We believe that this capability is broadly useful and applicable across product categories and industries. We share our learning to encourage the research and application of generative AI in different industries.