Google Research scientist Zhengdao Chen presented a paper at ICLR 2026 showing that diffusion models' creativity arises from neural networks learning a smoothed score function, which causes the model …
Four months after Google Research and Google DeepMind published TurboQuant, a training-free algorithm suite claiming 6x memory reduction and 8x speedup in attention computation on H100s, the company h…
Researchers reported on July 8, 2026, that AI peer-review systems can be manipulated by rewriting papers to satisfy model-generated reviewer preferences rather than improving scientific quality. Tests…
KakaoBank's Financial Technology Research Institute had four papers accepted at major AI conferences in 2026, including a prompt injection detection method at ICLR 2026 and studies at LREC 2026 and AC…
Researchers introduced PCB-Bench, the first comprehensive benchmark for evaluating large language models on printed circuit board placement and routing tasks. The benchmark includes over 3,700 questio…
Sakana AI launched Fugu, a multi-agent system delivered as a single model that dynamically orchestrates frontier models to tackle complex multi-step tasks. The system is accessible via the Sakana API …
Google Research developed MoGen, an AI model that generates synthetic neuron geometries to improve brain mapping. The model reduced reconstruction errors by 4.4%, potentially saving 157 person-years o…
CEM888.AI announced that its agent Vetta achieved the highest published scores on the MemoryAgentBench benchmark at ICLR 2026, with 99.9% on AR Retrieval and 77.2% on BEAM Memory, outperforming GPT-4.…
Researchers have developed DiffusionBlocks, a method that trains neural networks by dividing them into blocks and training only one block at a time, reducing memory requirements to roughly 1/B of stan…