Search for: All records

AI-generated images have become pervasive, raising critical concerns around content authenticity, intellectual property, and the spread of misinformation. Invisible watermarks offer a promising solution for identifying AI-generated images, preserving content provenance without degrading visual quality. However, their real-world robustness remains uncertain due to the lack of standardized evaluation protocols and large-scale stress testing. To bridge this gap, we organized “Erasing the Invisible,” a NeurIPS 2024 competition and newly established benchmark designed to systematically stress testing the resilience of watermarking techniques. The competition introduced two attack tracks—Black-box and Beige-box—that simulate practical scenarios with varying levels of attacker knowledge on watermarks, providing a comprehensive assessment of watermark robustness. The competition attracted significant global participation, with 2,722 submissions from 298 teams. Through a rigorous evaluation pipeline featuring real-time feedback and human-verified final rankings, participants developed and demonstrated new attack strategies that revealed critical vulnerabilities in state-of-the-art watermarking methods. On average, the top-5 teams in both tracks could remove watermarks from $$\geq$$ 89% of the images while preserving high visual quality, setting strong baselines for future research on watermark attacks and defenses. To support continued progress in this field, we summarize the insights and lessons learned from this competition in this paper, and release the benchmark dataset, evaluation toolkit, and competition results. “Erasing the Invisible” establishes a valuable open resource for advancing more robust watermarking techniques and strengthening content provenance in the era of generative AI. 

https://arxiv.org/pdf/2503.02951? 2025-acl-kodcode We introduce KodCode, a synthetic dataset that addresses the persistent challenge of acquiring high-quality, verifiable training data across diverse difficulties and domains for training Large Language Models for coding. Existing code-focused resources typically fail to ensure either the breadth of coverage (e.g., spanning simple coding tasks to advanced algorithmic problems) or verifiable correctness (e.g., unit tests). In contrast, KodCode comprises question-solution-test triplets that are systematically validated via a self-verification procedure. Our pipeline begins by synthesizing a broad range of coding questions, then generates solutions and test cases with additional attempts allocated to challenging problems. Finally, post-training data synthesis is done by rewriting questions into diverse formats and generating responses under a test-based reject sampling procedure from a reasoning model (DeepSeek R1). This pipeline yields a large-scale, robust and diverse coding dataset. KodCode is suitable for supervised fine-tuning and the paired unit tests also provide great potential for RL tuning. Fine-tuning experiments on coding benchmarks (HumanEval(+), MBPP(+), BigCodeBench, and LiveCodeBench) demonstrate that KodCode-tuned models achieve state-of-the-art performance, surpassing models like Qwen2.5-Coder-32B-Instruct and DeepSeek-R1-Distill-Llama-70B. 

St Helena Bay (SHB), a retentive zone in the productive southern Benguela Upwelling System off western South Africa, experiences seasonal hypoxia and episodic anoxic events that threaten local fisheries. To understand the drivers of oxygen variability in SHB, we queried 25 years of dissolved oxygen (DO) observations alongside high‐resolution wind and hydrographic data, and dynamical data from a high‐resolution model. At 70 m in SHB (mid‐bay), upwelling‐favorable winds in spring drove replenishment of cold, oxygenated water. Hypoxia developed in summer, becoming most severe in autumn. Bottom waters in autumn were replenished with warmer, less oxygenated water than in spring—suggesting a seasonal change in source waters upwelled into the bay. Downwelling and deep mixing in winter ventilated mid‐bay bottom waters, which reverted to hypoxic conditions during wind relaxations and reversals. In the nearshore (20 m), hypoxia occurred specifically during periods of upwelling‐favorable wind stress and was most severe in autumn. Using a statistical model, we extended basic hydrographic observations to nitrate and DO concentrations and developed metrics to identify the accumulation of excess nutrients on the shelf and nitrogen‐loss to denitrification, both of which were most prominent in autumn. A correspondence of the biogeochemical properties of hypoxic waters at 20 m to those at 70 m implicates the latter as the source waters upwelled inshore in autumn. We conclude that wind‐driven upwelling drives the replenishment of respired bottom waters in SHB with oxygenated waters, noting that less‐oxygenated water is imported later in the upwelling season, which exacerbates hypoxia.