Benchmarks are essential for quantitatively tracking progress in AI. As AI agents become increasingly capable, researchers and practitioners have introduced agentic benchmarks to evaluate agents on complex, real-world tasks. These benchmarks typically measure agent capabilities by evaluating task outcomes via specific reward designs. However, we show that many agentic benchmarks have issues in task setup or reward design. For example, SWE-bench Verified uses insufficient test cases, while TAU-bench counts empty responses as successful. Such issues can lead to under- or overestimation of agents' performance by up to 100% in relative terms. To make agentic evaluation rigorous, we introduce the Agentic Benchmark Checklist (ABC), a set of guidelines that we synthesized from our benchmark-building experience, a survey of best practices, and previously reported issues. When applied to CVE-Bench, a benchmark with a particularly complex evaluation design, ABC reduces the performance overestimation by 33%.

Nonverbal communication (NVC) plays an integral role in human language, but studying NVC in general is challenging because of its broad scope and high variance in interpretation among individuals and cultures. However, mime -- the theatrical technique of suggesting intent using only gesture, expression, and movement -- is a subset of NVC that consists of explicit and embodied actions with much lower human interpretation variance. We argue that a solid understanding of mimed actions is a crucial prerequisite for vision-language models capable of interpreting and commanding more subtle aspects of NVC. Hence, we propose Mime Identification Multimodal Evaluation (MIME), a novel video-based question answering benchmark comprising of 86 mimed actions. Constructed with motion capture data, MIME consists of variations of each action with perturbations applied to the character, background, and viewpoint for evaluating recognition robustness. We find that both open-weight and API-based vision-language models perform significantly worse than humans on MIME, motivating the need for increased research for instilling more robust understanding of human gestures.

Large Language Models (LLMs) offer new potential for automating documentation-to-code traceability, yet their capabilities remain underexplored. We present a comprehensive evaluation of LLMs (Claude 3.5 Sonnet, GPT-4o, and o3-mini) in establishing trace links between various software documentation (including API references and user guides) and source code. We create two novel datasets from two open-source projects (Unity Catalog and Crawl4AI). Through systematic experiments, we assess three key capabilities: (1) trace link identification accuracy, (2) relationship explanation quality, and (3) multi-step chain reconstruction. Results show that the best-performing LLM achieves F1-scores of 79.4% and 80.4% across the two datasets, substantially outperforming our baselines (TF-IDF, BM25, and CodeBERT). While fully correct relationship explanations range from 42.9% to 71.1%, partial accuracy exceeds 97%, indicating that fundamental connections are rarely missed. For multi-step chains, LLMs maintain high endpoint accuracy but vary in capturing precise intermediate links. Error analysis reveals that many false positives stem from naming-based assumptions, phantom links, or overgeneralization of architectural patterns. We demonstrate that task-framing, such as a one-to-many matching strategy, is critical for performance. These findings position LLMs as powerful assistants for trace discovery, but their limitations could necessitate human-in-the-loop tool design and highlight specific error patterns for future research.

--Large Language Models (LLMs) have significant impact on the healthcare scenarios but remain prohibitively large for deployment in real-time, resource-constrained environments such as edge devices. In this work, we introduce a novel medical assistant system, optimized through our general-purpose compression framework, which tailors Large Language Models (LLMs) for deployment in specialized domains. By measuring neuron saliency on domain-specific data, our method can aggressively prune irrelevant neurons, reducing model size while preserving performance. Following pruning, we apply post-training quantization to further reduce the memory footprint, and evaluate the compressed model across medical benchmarks including MedMCQA, MedQA, and PubMedQA. We also deploy the 50% compressed Gemma and the 67% compressed LLaMA3 models on Jetson Orin Nano (18.7W peak) and Raspberry Pi 5 (6.3W peak), achieving real-time, energy-efficient inference under hardware constraints.

In this paper, we present Optimized Prompt-based Unified System (OPUS), a framework that utilizes a Large Language Model (LLM) to control Pan-Tilt-Zoom (PTZ) cameras, providing contextual understanding of natural environments. To achieve this goal, the OPUS system improves cost-effectiveness by generating keywords from a high-level camera control API and transferring knowledge from larger closed-source language models to smaller ones through Supervised Fine-Tuning (SFT) on synthetic data. This enables efficient edge deployment while maintaining performance comparable to larger models like GPT-4. OPUS enhances environmental awareness by converting data from multiple cameras into textual descriptions for language models, eliminating the need for specialized sensory tokens. In benchmark testing, our approach significantly outperformed both traditional language model techniques and more complex prompting methods, achieving a 35% improvement over advanced techniques and a 20% higher task accuracy compared to closed-source models like Gemini Pro. The system demonstrates OPUS's capability to simplify PTZ camera operations through an intuitive natural language interface. This approach eliminates the need for explicit programming and provides a conversational method for interacting with camera systems, representing a significant advancement in how users can control and utilize PTZ camera technology.

The primary goal of this study is to develop and evaluate an innovative prompting technique, AnaQuest, for generating multiple-choice questions (MCQs) using a pre-trained large language model. In AnaQuest, the choice items are sentence-level assertions about complex concepts. The technique integrates formative and summative assessments. In the formative phase, students answer open-ended questions for target concepts in free text. For summative assessment, AnaQuest analyzes these responses to generate both correct and incorrect assertions. To evaluate the validity of the generated MCQs, Item Response Theory (IRT) was applied to compare item characteristics between MCQs generated by AnaQuest, a baseline ChatGPT prompt, and human-crafted items. An empirical study found that expert instructors rated MCQs generated by both AI models to be as valid as those created by human instructors. However, IRT-based analysis revealed that AnaQuest-generated questions - particularly those with incorrect assertions (foils) - more closely resembled human-crafted items in terms of difficulty and discrimination than those produced by ChatGPT.

Recent advancements in explainable recommendation have greatly bolstered user experience by elucidating the decision-making rationale. However, the existing methods actually fail to provide effective feedback signals for potentially better or worse generated explanations due to their reliance on traditional supervised learning paradigms in sparse interaction data. To address these issues, we propose a novel human-like feedback-driven optimization framework. This framework employs a dynamic interactive optimization mechanism for achieving human-centered explainable requirements without incurring high labor costs. Specifically, we propose to utilize large language models (LLMs) as human simulators to predict human-like feedback for guiding the learning process. To enable the LLMs to deeply understand the task essence and meet user's diverse personalized requirements, we introduce a human-induced customized reward scoring method, which helps stimulate the language understanding and logical reasoning capabilities of LLMs. Furthermore, considering the potential conflicts between different perspectives of explanation quality, we introduce a principled Pareto optimization that transforms the multi-perspective quality enhancement task into a multi-objective optimization problem for improving explanation performance. At last, to achieve efficient model training, we design an off-policy optimization pipeline. By incorporating a replay buffer and addressing the data distribution biases, we can effectively improve data utilization and enhance model generality. Extensive experiments on four datasets demonstrate the superiority of our approach.

Language representation learning has emerged as a promising approach for sequential recommendation, thanks to its ability to learn generalizable representations. However, despite its advantages, this approach still struggles with data sparsity and a limited understanding of common-sense user preferences. To address these limitations, we propose $\textbf{JEPA4Rec}$, a framework that combines $\textbf{J}$oint $\textbf{E}$mbedding $\textbf{P}$redictive $\textbf{A}$rchitecture with language modeling of item textual descriptions. JEPA4Rec captures semantically rich and transferable representations, improving recommendation performance and reducing reliance on large-scale pre-training data. Specifically, JEPA4Rec represents items as text sentences by flattening descriptive information such as $\textit{title, category}$, and other attributes. To encode these sentences, we employ a bidirectional Transformer encoder with modified embedding layers tailored for capturing item information in recommendation datasets. We apply masking to text sentences and use them to predict the representations of the unmasked sentences, helping the model learn generalizable item embeddings. To further improve recommendation performance and language understanding, we employ a two-stage training strategy incorporating self-supervised learning losses. Experiments on six real-world datasets demonstrate that JEPA4Rec consistently outperforms state-of-the-art methods, particularly in cross-domain, cross-platform, and low-resource scenarios.

Yesterday evening, Sam Altman shared an image of the Death Star on X. There was no caption on the picture, which showed the world-destroying Star Wars space station rising over an Earth-like planet, but his audience understood the context. In fewer than 24 hours, OpenAI would release an AI model intended to wipe out all the rest. That model, GPT-5, launched earlier today with all the requisite fanfare. In an announcement video, Altman said that the product will serve as a "legitimate Ph.D.-level expert in anything--any area you need, on demand--that can help you with whatever your goals are."

NVIDIA CEO and co-founder Jensen Huang commends President Donald Trump's A.I. agenda and outlines what the country's job future will look like on'Special Report.' Three years ago, I said Google was going the way of the dial-up modem. People called me crazy with a capital K. Well, I was spot on. ChatGPT now has over 180 million users and powers more than 800 million sessions each week. Google's own AI Overviews appear in over 60% of search results.