OpenAI Open Sources BrowseComp: A New Benchmark for Measuring the Ability for AI Agents to Browse the Web Despite advances in large language models (LLMs), AI agents still face notable limitations when navigating the open web to retrieve complex information. While many models excel on static knowledge benchmarks, they often underperform when tasked with locating… →

Google AI Introduces Ironwood: A Google TPU Purpose-Built for the Age of Inference At the 2025 Google Cloud Next event, Google introduced Ironwood , its latest generation of Tensor Processing Units (TPUs), designed specifically for large-scale AI inference workloads. This release marks a strategic shift toward optimizing infrastructure for inference, reflecting the increasing operational focus… →

ByteDance Introduces VAPO: A Novel Reinforcement Learning Framework for Advanced Reasoning Tasks In the Large Language Models (LLM) RL training, value-free methods like GRPO and DAPO have shown great effectiveness. The true potential lies in value-based methods, which allow more precise credit assignment by accurately tracing each action& impact on subsequent returns. This precision is… →

T* and LV-Haystack: A Spatially-Guided Temporal Search Framework for Efficient Long-Form Video Understanding Understanding long-form videos—ranging from minutes to hours—presents a major challenge in computer vision, especially as video understanding tasks expand beyond short clips. One of the key difficulties lies in efficiently identifying the few relevant frames from thousands within a lengthy video necessary… →

This AI Paper Introduces a Machine Learning Framework to Estimate the Inference Budget for Self-Consistency and GenRMs (Generative Reward Models) Large Language Models (LLMs) have demonstrated significant advancements in reasoning capabilities across diverse domains, including mathematics and science. However, improving these reasoning abilities at test time remains a challenge researchers are actively addressing. The primary… →