#AI research

OpenAI has released its first open-weight large language models since GPT-2, offering downloadable models under a permissive license that support customization and local use, marking a strategic move in AI research and geopolitics.

OpenAI is pushing the boundaries of AI by focusing on human-like reasoning and creativity, highlighted by recent successes in coding and math competitions and their ongoing AGI research.

Discover how context engineering advances large language models beyond prompt engineering with innovative techniques, system architectures, and future research directions.

Google introduces TTD-DR, an innovative AI framework that mimics human research processes through iterative draft refinement, achieving top performance in complex research benchmarks.

Explore the roles of Mark Chen and Jakub Pachocki in driving OpenAI's advanced research and the development of AI models like GPT-5, highlighting recent achievements and challenges in the race for AGI.

Alibaba introduces Lumos-1, a unified autoregressive video generation model leveraging MM-RoPE and AR-DF to efficiently model spatiotemporal dependencies, achieving competitive results with efficient training.

GLM-4.1V-Thinking is a cutting-edge vision-language model that pushes the boundaries of multimodal reasoning, setting new standards across various challenging AI tasks.

Mirage introduces a new method for vision-language models to integrate visual reasoning without generating images, significantly enhancing their ability to solve spatial and multimodal tasks.

MMSearch-R1 introduces a reinforcement learning framework that enables large multimodal models to perform efficient, on-demand searches by learning when and how to retrieve relevant information, significantly improving accuracy and reducing search overhead.

'ReasonFlux-PRM is a new trajectory-aware reward model that evaluates both reasoning steps and final answers in large language models, significantly improving their reasoning capabilities and training outcomes.'

GURU introduces a multi-domain reinforcement learning dataset and models that significantly improve reasoning abilities of large language models across six diverse domains, outperforming previous open models.

ByteDance introduces VGR, a multimodal large language model that dynamically integrates visual information during reasoning, significantly improving accuracy and efficiency in vision-language benchmarks.

ByteDance researchers introduce ProtoReasoning, a new framework leveraging logic-based prototypes to significantly improve reasoning and planning abilities in large language models across various domains.

Internal Coherence Maximization (ICM) introduces a novel label-free, unsupervised training framework for large language models, achieving performance on par with human-supervised methods and enabling advanced capabilities without human feedback.

Discover how to create a flexible asynchronous AI agent network with Google’s Gemini models, enabling collaborative research, analysis, and validation through specialized agent roles.

'Researchers introduce Soft Thinking, a training-free method that allows large language models to reason with continuous concept embeddings, enhancing accuracy and efficiency in math and coding tasks.'

Researchers improve large language models' reasoning by explicitly aligning core abilities like deduction, induction, and abduction, surpassing traditional instruction-tuned models in accuracy and reliability.

Salesforce Research introduces UAEval4RAG, a new benchmark framework that evaluates RAG systems' ability to reject unanswerable queries across diverse categories, enhancing the reliability of AI responses.

Google researchers introduce LightLab, a novel diffusion-based AI technique that enables fine-grained, physically plausible control of lighting in single images, outperforming existing methods.

ByteDance introduces Seed1.5-VL, a powerful vision-language model achieving state-of-the-art performance on numerous benchmarks, advancing multimodal AI understanding and reasoning.

'Alibaba’s ZeroSearch framework leverages reinforcement learning and simulated document generation to train language models for retrieval without relying on costly real-time search APIs, achieving performance comparable to or better than Google Search.'

WebThinker is a new AI agent that empowers large reasoning models to autonomously search the web and generate detailed scientific reports, significantly improving performance on complex reasoning benchmarks.

NVIDIA, CMU, and Boston University researchers introduce Nemotron-CrossThink, a novel framework that expands reinforcement learning for large language models beyond math to multiple reasoning domains with improved accuracy and efficiency.