A team led by Ming Zhong at the University of Illinois Urbana-Champaign and Google DeepMind has introduced SWE-IF, a framework that aligns code evaluation with human preference, revealing that instruc…
A first-time AI safety experiment by Theodore P. J. found that doubling a user profile changes the effect of an instruction about using saved memories, with results suggesting the instruction's impact…
Hugging Face reported that coding agents fully reproduced every extracted claim in 266 of 2,226 ICML 2026 papers, with 632 more showing partial evidence and no falsified claims. The 19-day challenge, …
A community hackathon reproducing 2,200 papers from ICML 2026 found that 51% (1,103) had at least one claim independently verified, while 23% (496) had at least one claim falsified or contested, accor…
An ICML 2026 paper submitted July 9 reports that amplifying the weight-space difference between Qwen3-VL instruct and reasoning checkpoints made planted secrets or unintended behaviors surface up to 1…
A researcher argues that large language models (LLMs) could achieve scientific breakthroughs like General Relativity through deductive reasoning, countering the ICML 2026 position paper by Tom Zahavy …
Together AI introduces ThunderAgent, a system for high-throughput agentic inference that achieves up to 2.5× higher single-node throughput and 2.4× speedup on an 8-node cluster with near-linear scalin…
Researchers at Sakana AI introduced UnMaskFork (UMF), a test-time scaling method that enables multiple masked diffusion language models to collaborate on a single answer, outperforming existing method…
Researchers from Zhejiang University and Alibaba presented a study at ICML 2026 showing that logically inconsistent prompts can cause reasoning models to enter long, unproductive internal loops, effec…
LG AI Research is showcasing real-world industrial applications of its EXAONE large language model at ICML 2026 in Seoul, including EXAONE Discovery for materials and drug development. The company hig…
Researchers introduced Sheaf-ADMM, a framework for multi-agent coordination that divides complex tasks into overlapping pieces assigned to individual agents. In tests, it achieved a 93% solve rate on …
Princeton University researchers will present 12 papers at the 43rd International Conference on Machine Learning (ICML 2026), held July 6-11 in Seoul, South Korea, including four oral presentations an…
Researchers at ICML 2026 presented a unified framework showing that parametric and nonparametric black-box optimizers are variations of a single update equation. By bridging this theoretical gap, they…
Microsoft researchers introduced Memora, a scalable memory system for AI agents that decouples stored content from retrieval mechanisms, achieving state-of-the-art performance on long-horizon benchmar…
Sean Xuefeng Du will present a new hallucination-detection method, Answer-agreement Representation Shaping (ARS), at ICML 2026 in Seoul on July 9. The work, developed at Nanyang Technological Universi…
Researchers at Georgia Tech's AMIR Lab have developed ProtoMech, a framework that traces internal computational circuits inside protein language models like ESM2, revealing that these AI systems indep…