Reinforcement Learning with Verifiable Rewards (RLVR) improves instruction following capabilities of large language models (LLMs), but suffers from training inefficiency due to inadequate difficulty assessment. Moreover, RLVR is prone to over-optimization, where LLMs exploit verification shortcuts without aligning to the actual intent of user instructions. We introduce Instruction Following Decorator (IFDecorator}, a framework that wraps RLVR training into a robust and sample-efficient pipeline. It consists of three components: (1) a cooperative-adversarial data flywheel that co-evolves instructions and hybrid verifications, generating progressively more challenging instruction-verification pairs; (2) IntentCheck, a bypass module enforcing intent alignment; and (3) trip wires, a diagnostic mechanism that detects reward hacking via trap instructions, which trigger and capture shortcut exploitation behaviors. Our Qwen2.5-32B-Instruct-IFDecorator achieves 87.43% accuracy on IFEval, outperforming larger proprietary models such as GPT-4o. Additionally, we demonstrate substantial improvements on FollowBench while preserving general capabilities. Our trip wires show significant reductions in reward hacking rates. We will release models, code, and data for future research.

Functional object arrangement (FORM) is the task of arranging objects to fulfill a function, e.g., "set up a dining table for two". One key challenge here is that the instructions for FORM are often under-specified and do not explicitly specify the desired object goal poses. This paper presents SetItUp, a neuro-symbolic framework that learns to specify the goal poses of objects from a few training examples and a structured natural-language task specification. SetItUp uses a grounding graph, which is composed of abstract spatial relations among objects (e.g., left-of), as its intermediate representation. This decomposes the FORM problem into two stages: (i) predicting this graph among objects and (ii) predicting object poses given the grounding graph. For (i), SetItUp leverages large language models (LLMs) to induce Python programs from a task specification and a few training examples. This program can be executed to generate grounding graphs in novel scenarios. For (ii), SetItUp pre-trains a collection of diffusion models to capture primitive spatial relations and online composes these models to predict object poses based on the grounding graph. We evaluated SetItUp on a dataset spanning three distinct task families: arranging tableware on a dining table, organizing items on a bookshelf, and laying out furniture in a bedroom. Experiments show that SetItUp outperforms existing models in generating functional, physically feasible, and aesthetically pleasing object arrangements. This article extends our conference paper published at Robotics: Science and Systems (RSS) 2024.

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%.

The focus of nowcasting development is transitioning from physically motivated advection methods to purely data-driven Machine Learning (ML) approaches. Nevertheless, recent work indicates that incorporating advection into the ML value chain has improved skill for radar-based precipitation nowcasts. However, the generality of this approach and the underlying causes remain unexplored. This study investigates the generality by probing the approach on satellite-based thunderstorm nowcasts for the first time. Resorting to a scale argument, we then put forth an explanation when and why skill improvements can be expected. In essence, advection guarantees that thunderstorm patterns relevant for nowcasting are contained in the receptive field at long forecast times. To test our hypotheses, we train ResU-Nets solving segmentation tasks with lightning observations as ground truth. The input of the Baseline Neural Network (BNN) are short time series of multispectral satellite imagery and lightning observations, whereas the Advection-Informed Neural Network (AINN) additionally receives the Lagrangian persistence nowcast of all input channels at the desired forecast time. Overall, we find only a minor skill improvement of the AINN over the BNN when considering fully averaged scores. However, assessing skill conditioned on forecast time and advection speed, we demonstrate that our scale argument correctly predicts the onset of skill improvement of the AINN over the BNN after 2h forecast time. We confirm that, generally, advection becomes gradually more important with longer forecast times and higher advection speeds. Our work accentuates the importance of considering and incorporating the underlying physical scales when designing ML-based forecasting models.

For tourists visiting South Korea, one of the world's most technologically advanced nations, navigating the country's urban heartlands can prove surprisingly frustrating for one simple reason: Google Maps just doesn't work effectively. But on 11 August that could change, as South Korean authorities are set to decide whether to finally grant Google's request to export the country's detailed mapping data to overseas servers. Such a move would open up functionality that allows the app to give detailed directions and show users the best routes to travel. It is a debate spanning nearly two decades which has evolved into a broader test of how democracies balance digital sovereignty with economic openness. Local industry groups are warning of market domination from foreign companies, while those who back Google's request argue restrictions harm tourism and innovation.

NASA's Curiosity rover snapped a bizarre, coral-shaped rock on the surface of Mars, sparking fresh speculation about signs of ancient life on the Red Planet. The twisted, alien-like formation was sculpted by wind and time, according to NASA, which said it likely formed billions of years ago when water once flowed across the Martian surface. The images have taken the internet by storm, with some users claiming: 'Corals are true signs of ancient life forms along with the ancient rivers. This is a huge discovery!!' Another wrote on X: 'There's your Mars fossilized foreign life material evidence everybody's been asking for. That's obviously been there all along.'

As anyone with even a trace of a regional dialect who has had to pay a parking fine can attest, voice recognition services struggle with accents. Now, people in Mark Sewards' constituency in Leeds are likely to find the same problem with his AI variant. A chatbot billed as the first AI version of an MP responds in Sewards' voice with advice, support or by offering to pass on a message to his team – but only if it understands you. The website, a virtual representation of the MP for Leeds South West and Morley – complete with a Pixar-style cartoon – was launched by a local startup to field questions from his constituents, some of whom have broad Leeds accents. I was interested to see how "Sewardsbot" would handle a conversation with someone from only a couple of miles away from his constituency border.

U.S. Ambassador to NATO Matthew Whitaker responds to the'reckless threatening' on'Fox & Friends First' amid a rush to transfer patriot missiles to Ukraine. Days ahead of the U.S. preparing harsh new sanctions tied to the war in Ukraine, Russian Vladimir Putin, whether intentionally or just carelessly, has tested the political will of NATO's collective defense guarantee, Article 5. In recent days, drones launched from the Russian-aligned state of Belarus have pierced Lithuanian airspace, drawing alarms from the region's political and military leaders. One drone traversed approximately 100 kilometers, loitered ominously over Vilnius carrying two kilograms of explosives and ultimately crashed inside a military training zone. Earlier in July, another drone forced the evacuation of high-level officials when it crashed near the Šumskas border crossing.

Conservative Gen Z influencer Bo Loudon and National Review staff writer Caroline Downey weigh in on the Sydney Sweeney jeans ad and President Donald Trump's support of her on'The Ingraham Angle.' President Donald Trump notched his 200th day back in office Thursday, with the administration celebrating a lengthy list of wins across its latest chapter of actions and policies unfolding at a breakneck pace. "In just 200 days, President Trump has turned America into the hottest country in the world," White House spokeswoman Taylor Rogers told Fox News Digital. "Under Joe Biden's failed leadership, families and businesses were struggling, and America was dead -- but President Trump has quickly restored American greatness. The historic trade deals and peace deals he secured on behalf of the American people made President Trump's second 100 days just as successful as the first." Trump hit his 100th day of his second administration in April, which included operating at warp speed as Trump signed dozens of executive orders, leveled harsh tariffs on foreign nations to bring parity to the U.S.' trade deficit, negotiated with foreign nations to work to end wars, unveiled the Department of Government Effeciency to investigate the federal government for potential mismanagement and fraud, locked down the U.S. border with Mexico and continued an overhaul of the federal government so it falls in line with the admin's "America First" policies.

United States President Donald Trump has fired off a social media message calling on the head of the US technology firm Intel to resign from his post as chief executive officer. Trump's decision to denounce Intel CEO Lip-Bu Tan on Thursday morning sent the company's stocks tumbling, amid the uncertainty about the future of its leadership. "The CEO of INTEL is highly CONFLICTED and must resign, immediately," Trump wrote. "There is no other solution to this problem. Thank you for your attention to this problem!" Trump's post appeared to be a response to reports that Tan has invested nearly 200m in Chinese technology manufacturing and chip firms, including some with links to the country's military.