Pruning large language models (LLMs) is a promising solution for reducing model sizes and computational complexity while preserving performance. Traditional layer-wise pruning methods often adopt a uniform sparsity approach across all layers, which leads to suboptimal performance due to the varying significance of individual transformer layers within the model not being accounted for. To this end, we propose the Shapley Value-based Non-Uniform Pruning (SV-NUP) method for LLMs. This approach quantifies the contribution of each transformer layer to the overall model performance, enabling the assignment of tailored pruning budgets to different layers to retain critical parameters. To further improve efficiency, we design the Sliding Window-based Shapley Value approximation method. It substantially reduces computational overhead compared to exact SV calculation methods. Extensive experiments on various LLMs including LLaMA-v1, LLaMA-v2 and OPT demonstrate the effectiveness of the proposed approach. The results reveal that non-uniform pruning significantly enhances the performance of pruned models. Notably, SV-NUP achieves a reduction in perplexity (PPL) of 18.01% and 19.55% on LLaMA-7B and LLaMA-13B, respectively, compared to SparseGPT at 70% sparsity.

Recent advances in reasoning-enhanced large language models (LLMs) and multimodal LLMs (MLLMs) have significantly improved performance in complex tasks, yet medical AI models often overlook the structured reasoning processes inherent in clinical practice. In this work, we present ChestX-Reasoner, a radiology diagnosis MLLM designed to leverage process supervision mined directly from clinical reports, reflecting the step-by-step reasoning followed by radiologists. We construct a large dataset by extracting and refining reasoning chains from routine radiology reports. Our two-stage training framework combines supervised fine-tuning and reinforcement learning guided by process rewards to better align model reasoning with clinical standards. We introduce RadRBench-CXR, a comprehensive benchmark featuring 59K visual question answering samples with 301K clinically validated reasoning steps, and propose RadRScore, a metric evaluating reasoning factuality, completeness, and effectiveness. ChestX-Reasoner outperforms existing medical and general-domain MLLMs in both diagnostic accuracy and reasoning ability, achieving 16%, 5.9%, and 18% improvements in reasoning ability compared to the best medical MLLM, the best general MLLM, and its base model, respectively, as well as 3.3%, 24%, and 27% improvements in outcome accuracy. All resources are open-sourced to facilitate further research in medical reasoning MLLMs.

It's the same but not the same: Do LLMs distinguish Spanish varieties? Abstract: In recent years, large language models (LLMs) have demonstrated a high capacity for understanding and generating text in Spanish. However, with five hundred million native speakers, Spanish is not a homogeneous language but rather one rich in diatopic vari ations spanning both sides of the Atlantic. For this reason, in this study, we evaluate the ability of nine language models to identify and distinguish the morphosyntactic and lexical peculiarities of seven varieties of Spanish (Andean, Antillean, Continen tal Caribbean, Chilean, Peninsular, Mexican and Central American and Rioplatense) through a multiple - choice test. The results indicate that the Peninsular Spanish variety is the best identified by all models and that, among them, GPT - 4o is the only model c apable of recognizing the variability of the Spanish language.

The improvement of LLMs' instruction-following capabilities depends critically on the availability of high-quality instruction-response pairs. While existing automatic data synthetic methods alleviate the burden of manual curation, they often rely heavily on either the quality of seed data or strong assumptions about the structure and content of web documents. To tackle these challenges, we propose Web Reconstruction (WebR), a fully automated framework for synthesizing high-quality instruction-tuning (IT) data directly from raw web documents with minimal assumptions. Leveraging the inherent diversity of raw web content, we conceptualize web reconstruction as an instruction-tuning data synthesis task via a novel dual-perspective paradigm--Web as Instruction and Web as Response--where each web document is designated as either an instruction or a response to trigger the reconstruction process. Comprehensive experiments show that datasets generated by WebR outperform state-of-the-art baselines by up to 16.65% across four instruction-following benchmarks. Notably, WebR demonstrates superior compatibility, data efficiency, and scalability, enabling enhanced domain adaptation with minimal effort. The data and code are publicly available at https://github.com/YJiangcm/WebR.

Recent work has demonstrated the remarkable potential of Large Language Models (LLMs) in test-time scaling. By making models think before answering, they are able to achieve much higher accuracy with extra inference computation. However, in many real-world scenarios, models are used under time constraints, where an answer should be given within a certain output length. It is unclear whether and how the reasoning ability of different LLMs remain effective under strict constraints. We take a first look at this problem by conducting an in-depth empirical study. Specifically, we test 30 LLMs on common reasoning datasets under a wide range of output length budgets, and we analyze the correlation between the inference accuracy and various properties including model type, model size, prompt style, etc. We also consider the mappings between token budgets and actual on-device latency budgets. The results have demonstrated several interesting findings regarding the budget-aware LLM reasoning ability that differ from the unconstrained situation, e.g. the optimal choices of either model size or prompt style change under different budgets. These findings offer timely evaluation to this area and practical guidance for users to deploy LLMs under real-world latency constraints.

The legendary designer behind Apple's iPhone, Jony Ive, has joined OpenAI to create devices tailored for using generative artificial intelligence, according to a video posted Wednesday by the ChatGPT maker. Ive and his team will take over design at OpenAI as part of an acquisition of his startup named "IO" valued at 6.5 billion. Sharing no details, OpenAI chief executive Sam Altman said in the video that a prototype Ive shared with him "is the coolest piece of technology that the world will have ever seen."

OpenAI is buying an untested startup for 6.4bn, the ChatGPT maker's biggest acquisition yet. The hardware startup, called io, was founded by Apple design guru Jony Ive, known best as one of the principal architects of the iPhone. Ive and OpenAI's CEO, Sam Altman, said in a blog post that their partnership has been two years in the making. "A collaboration built upon friendship, curiosity and shared values quickly grew in ambition," they wrote in the blog post, which offered scant details on upcoming devices. "Tentative ideas and explorations evolved into tangible designs."

OpenAI has fully acquired Io, a joint venture it cocreated last year with Jony Ive, the famed British designer behind the sleek industrial aesthetic that defined the iPhone and more than two decades of Apple products. In a nearly 10-minute video posted to X on Wednesday, Ive and OpenAI CEO Sam Altman said the Apple pioneer's "creative collective" will "merge with OpenAI to work more intimately with the research, engineering, and product teams in San Francisco." OpenAI says it's paying 5 billion in equity to acquire Io. The promotional video included musings on technology from both Ive and Altman, set against the golden-hour backdrop of the streets of San Francisco, but the two never share exactly what it is they're building. "We look forward to sharing our work next year," a text statement at the end of the video reads.

Sign up for the Slatest to get the most insightful analysis, criticism, and advice out there, delivered to your inbox daily. Over the past week, at least two venerable American newspapers--the Chicago Sun-Times and the Philadelphia Inquirer--published a 56-page insert of summer content that was in large part produced by A.I. The most glaring evidence was a now-notorious "summer reading list," which recommended 15 books, five of them real, 10 of them imaginary, with summaries of fake titles like Isabel Allende's Tidewater Dreams, Min Jin Lee's Nightshade Market, Rebecca Makkai's Boiling Point, and Percival Everett's The Rainmakers. The authors exist; the books do not. The rest of the section, which included anodyne listicles about summer activities, barbecuing, and photography, soon attracted additional scrutiny.

Large language models such as GPT, Llama, Claude, and DeepSeek can be so fluent that people feel it as a "you," and it answers encouragingly as an "I." The models can write poetry in nearly any given form, read a set of political speeches and promptly sift out and share all the jokes, draw a chart, code a website. How do they do these and so many other things that were just recently the sole realm of humans? Practitioners are left explaining jaw-dropping conversational rabbit-from-a-hat extractions with arm-waving that the models are just predicting one word at a time from an unthinkably large training set scraped from every recorded written or spoken human utterance that can be found--fair enough--or a with a small shrug and a cryptic utterance of "fine-tuning" or "transformers!" These aren't very satisfying answers for how these models can converse so intelligently, and how they sometimes err so weirdly.