Multimodal punchlines, which involve humor or sarcasm conveyed in image-caption pairs, are a popular way of communication on online multimedia platforms. With the rapid development of multimodal large language models (MLLMs), it is essential to assess their ability to effectively comprehend these punchlines. However, existing benchmarks on punchline comprehension suffer from three major limitations: 1) language shortcuts that allow models to solely rely on text, 2) lack of question diversity, and 3) narrow focus on a specific domain of multimodal content (e.g., cartoon). To address these limitations, we introduce a multimodal \textbf{Punch}line comprehension \textbf{Bench}mark, named \textbf{PunchBench}, which is tailored for accurate and comprehensive evaluation of punchline comprehension. To enhance the evaluation accuracy, we generate synonymous and antonymous captions by modifying original captions, which mitigates the impact of shortcuts in the captions. To provide a comprehensive evaluation, PunchBench incorporates diverse question formats and image-captions from various domains. On this basis, we conduct extensive evaluations and reveal a significant gap between state-of-the-art MLLMs and humans in punchline comprehension. To improve punchline comprehension, we propose Simple-to-Complex Chain-of-Question (SC-CoQ) strategy, enabling the models to incrementally address complicated questions by first mastering simple ones. SC-CoQ effectively enhances the performance of various MLLMs on PunchBench, surpassing in-context learning and chain-of-thought.

Interpretability for Table Question Answering (Table QA) is critical, particularly in high-stakes industries like finance or healthcare. Although recent approaches using Large Language Models (LLMs) have significantly improved Table QA performance, their explanations for how the answers are generated are ambiguous. To fill this gap, we introduce Plan-of-SQLs ( or POS), an interpretable, effective, and efficient approach to Table QA that answers an input query solely with SQL executions. Through qualitative and quantitative evaluations with human and LLM judges, we show that POS is most preferred among explanation methods, helps human users understand model decision boundaries, and facilitates model success and error identification. Furthermore, when evaluated in standard benchmarks (TabFact, WikiTQ, and FetaQA), POS achieves competitive or superior accuracy compared to existing methods, while maintaining greater efficiency by requiring significantly fewer LLM calls and database queries.

Variations in languages across geographic regions or cultures are crucial to address to avoid biases in NLP systems designed for culturally sensitive tasks, such as hate speech detection or dialog with conversational agents. In languages such as Spanish, where varieties can significantly overlap, many examples can be valid across them, which we refer to as common examples. Ignoring these examples may cause misclassifications, reducing model accuracy and fairness. Therefore, accounting for these common examples is essential to improve the robustness and representativeness of NLP systems trained on such data. In this work, we address this problem in the context of Spanish varieties. We use training dynamics to automatically detect common examples or errors in existing Spanish datasets. We demonstrate the efficacy of using predicted label confidence for our Datamaps \cite{swayamdipta-etal-2020-dataset} implementation for the identification of hard-to-classify examples, especially common examples, enhancing model performance in variety identification tasks. Additionally, we introduce a Cuban Spanish Variety Identification dataset with common examples annotations developed to facilitate more accurate detection of Cuban and Caribbean Spanish varieties. To our knowledge, this is the first dataset focused on identifying the Cuban, or any other Caribbean, Spanish variety.

The Myers-Briggs Type Indicator (MBTI) is one of the most influential personality theories reflecting individual differences in thinking, feeling, and behaving. MBTI personality detection has garnered considerable research interest and has evolved significantly over the years. However, this task tends to be overly optimistic, as it currently does not align well with the natural distribution of population personality traits. Specifically, (1) the self-reported labels in existing datasets result in incorrect labeling issues, and (2) the hard labels fail to capture the full range of population personality distributions. In this paper, we optimize the task by constructing MBTIBench, the first manually annotated high-quality MBTI personality detection dataset with soft labels, under the guidance of psychologists. As for the first challenge, MBTIBench effectively solves the incorrect labeling issues, which account for 29.58% of the data. As for the second challenge, we estimate soft labels by deriving the polarity tendency of samples. The obtained soft labels confirm that there are more people with non-extreme personality traits. Experimental results not only highlight the polarized predictions and biases in LLMs as key directions for future research, but also confirm that soft labels can provide more benefits to other psychological tasks than hard labels. The code and data are available at https://github.com/Personality-NLP/MbtiBench.

Language models (LMs) have shown outstanding performance in text summarization including sensitive domains such as medicine and law. In these settings, it is important that personally identifying information (PII) included in the source document should not leak in the summary. Prior efforts have mostly focused on studying how LMs may inadvertently elicit PII from training data. However, to what extent LMs can provide privacy-preserving summaries given a non-private source document remains under-explored. In this paper, we perform a comprehensive study across two closed- and three open-weight LMs of different sizes and families. We experiment with prompting and fine-tuning strategies for privacy-preservation across a range of summarization datasets across three domains. Our extensive quantitative and qualitative analysis including human evaluation shows that LMs often cannot prevent PII leakage on their summaries and that current widely-used metrics cannot capture context dependent privacy risks.

Transformer-based Large Language Models (LLMs) have exhibited remarkable success in various natural language processing tasks primarily attributed to selfattention mechanism, which requires a token to consider all preceding tokens as its context to compute the attention score. However, when the context length L becomes very large (e.g., 32K), more redundant context information will be included w.r.t. L; 2) The presence of redundant context information may hamper the model to capture dependencies among crucial tokens, which may degrade the representation performance. In this paper, we propose a plug-and-play Core Context Aware (CCA) Attention for efficient long-range context modeling, which consists of two components: 1) Globality-pooling attention that divides input tokens into groups and then dynamically merges tokens within each group into one core token based on their significance; 2) Locality-preserved attention that incorporates neighboring tokens into the attention calculation. The two complementary attentions will then be fused to the final attention, maintaining comprehensive modeling ability as the full self-attention. In this way, the core context information w.r.t. a given token will be automatically focused and strengthened, while the context information in redundant groups will be diminished during the learning process. As a result, the computational and memory complexity will be significantly reduced. More importantly, the CCA-Attention can improve the long-context modeling ability by diminishing the redundant context information. Extensive experimental results demonstrate that our CCA-Attention significantly outperforms state-of-theart models in terms of computational efficiency and long-context modeling ability. Large language models (LLMs) (Brown et al., 2020; OpenAI, 2023; Touvron et al., 2023a) have demonstrated exceptional proficiency across various applications by effectively modeling extended contexts, particularly in tasks involving natural language understanding and generation (Ouyang et al., 2022; Chang et al., 2024). The remarkable success of LLMs is predominantly credited to the self-attention mechanism (Vaswani et al., 2017), which requires each token in the input sequence to calculate attention with all preceding tokens. Nonetheless, the computational and memory requirements of self-attention grow quadratically with the increase in sequence length, posing challenges for long context understanding tasks (Liu et al., 2024; Shaham et al., 2023). Recent studies (Beltagy et al., 2020; Zaheer et al., 2020; Xiao et al., 2024b) have demonstrated that the majority of layers within autoregressive LLMs exhibit redundant tokens across various attention heads and input tokens. This redundancy is visually exemplified in Figure 1, where we present a detailed visualization of the attention weights within LLaMA2 (Touvron et al., 2023a).

The proliferation of Internet services has led to an increasing need to protect private data. User authentication serves as a crucial mechanism to ensure data security. Although robust authentication forms the cornerstone of remote service security, it can still leave users vulnerable to credential disclosure, device-theft attacks, session hijacking, and inadequate adaptive security measures. Risk-based Authentication (RBA) emerges as a potential solution, offering a multi-level authentication approach that enhances user experience without compromising security. In this paper, we propose a Federated Risk-based Authentication (F-RBA) framework that leverages Federated Learning to ensure privacy-centric training, keeping user data local while distributing learning across devices. Whereas traditional approaches rely on centralized storage, F-RBA introduces a distributed architecture where risk assessment occurs locally on users' devices. The framework's core innovation lies in its similarity-based feature engineering approach, which addresses the heterogeneous data challenges inherent in federated settings, a significant advancement for distributed authentication. By facilitating real-time risk evaluation across devices while maintaining unified user profiles, F-RBA achieves a balance between data protection, security, and scalability. Through its federated approach, F-RBA addresses the cold-start challenge in risk model creation, enabling swift adaptation to new users without compromising security. Empirical evaluation using a real-world multi-user dataset demonstrates the framework's effectiveness, achieving a superior true positive rate for detecting suspicious logins compared to conventional unsupervised anomaly detection models. This research introduces a new paradigm for privacy-focused RBA in distributed digital environments, facilitating advancements in federated security systems.

This paper presents the \textbf{S}emantic-a\textbf{W}ar\textbf{E} spatial-t\textbf{E}mporal \textbf{T}okenizer (SweetTokenizer), a compact yet effective discretization approach for vision data. Our goal is to boost tokenizers' compression ratio while maintaining reconstruction fidelity in the VQ-VAE paradigm. Firstly, to obtain compact latent representations, we decouple images or videos into spatial-temporal dimensions, translating visual information into learnable querying spatial and temporal tokens through a \textbf{C}ross-attention \textbf{Q}uery \textbf{A}uto\textbf{E}ncoder (CQAE). Secondly, to complement visual information during compression, we quantize these tokens via a specialized codebook derived from off-the-shelf LLM embeddings to leverage the rich semantics from language modality. Finally, to enhance training stability and convergence, we also introduce a curriculum learning strategy, which proves critical for effective discrete visual representation learning. SweetTokenizer achieves comparable video reconstruction fidelity with only \textbf{25\%} of the tokens used in previous state-of-the-art video tokenizers, and boost video generation results by \textbf{32.9\%} w.r.t gFVD. When using the same token number, we significantly improves video and image reconstruction results by \textbf{57.1\%} w.r.t rFVD on UCF-101 and \textbf{37.2\%} w.r.t rFID on ImageNet-1K. Additionally, the compressed tokens are imbued with semantic information, enabling few-shot recognition capabilities powered by LLMs in downstream applications.

Learning the Individual Treatment Effect (ITE) is essential for personalized decision making, yet causal inference has traditionally focused on aggregated treatment effects. While integrating conformal prediction with causal inference can provide valid uncertainty quantification for ITEs, the resulting prediction intervals are often excessively wide, limiting their practical utility. To address this limitation, we introduce \underline{S}urrogate-assisted \underline{C}onformal \underline{I}nference for \underline{E}fficient I\underline{N}dividual \underline{C}ausal \underline{E}ffects (SCIENCE), a framework designed to construct more efficient prediction intervals for ITEs. SCIENCE applies to various data configurations, including semi-supervised and surrogate-assisted semi-supervised learning. It accommodates covariate shifts between source data, which contain primary outcomes, and target data, which may include only surrogate outcomes or covariates. Leveraging semi-parametric efficiency theory, SCIENCE produces rate double-robust prediction intervals under mild rate convergence conditions, permitting the use of flexible non-parametric models to estimate nuisance functions. We quantify efficiency gains by comparing semi-parametric efficiency bounds with and without the incorporation of surrogates. Simulation studies demonstrate that our surrogate-assisted intervals offer substantial efficiency improvements over existing methods while maintaining valid group-conditional coverage. Applied to the phase 3 Moderna COVE COVID-19 vaccine trial, SCIENCE illustrates how multiple surrogate markers can be leveraged to generate more efficient prediction intervals.

Foundation models (FMs) are pre-trained on large-scale datasets and then fine-tuned on a downstream task for a specific application. The most successful and most commonly used fine-tuning method is to update the pre-trained weights via a low-rank adaptation (LoRA). LoRA introduces new weight matrices that are usually initialized at random with a uniform rank distribution across the model weights. Recent works focus on different initialization schemes or the learning of adaptive ranks during fine-tuning. Both approaches have only been investigated in isolation, resulting in slow convergence or a uniform rank distribution, in turn leading to suboptimal performance. We propose to improve LoRA by initializing the new weights in a data-driven manner by computing singular value decomposition (SVD) on minibatches of activation vectors. Then, we initialize the LoRA matrices with the obtained right-singular vectors and redistribute ranks among all weight matrices to provably store the maximum amount of information of the downstream data in the newly introduced weights. In this way, only what information to maintain or neglect during the fine-tuning process needs to be learned. We call our new method $\textbf{E}$xplained $\textbf{V}$ariance $\textbf{A}$daptation (EVA). We apply EVA to a variety of fine-tuning tasks ranging from language generation and understanding to image classification and reinforcement learning. EVA exhibits faster convergence than competitors and achieves the highest average score across a multitude of tasks per domain while reducing the number of trainable parameters through rank redistribution.