Records found: 21
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.
Explore the comprehensive 7-layer framework essential for building real-world autonomous AI agents capable of thinking, acting, and learning effectively in 2025.
Explore how to create a smart conversational AI agent with memory by integrating Cognee and free Hugging Face models. This tutorial covers setup, learning, reasoning, and conversation capabilities.
Energy-Based Transformers enable machines to perform advanced, unsupervised System 2 Thinking, improving reasoning and generalization across tasks and modalities without domain-specific supervision.
Discover how to use Mirascope and Groq’s LLaMA 3 model to implement Chain-of-Thought reasoning, enabling AI to solve complex problems step-by-step effectively.
MetaStone-S1 introduces a unified reflective generative approach that achieves OpenAI o3-mini-level reasoning performance with significantly reduced computational resources, pioneering efficient AI reasoning architectures.
AbstRaL uses reinforcement learning to teach LLMs abstract reasoning, significantly improving their robustness and accuracy on varied GSM8K math problems compared to traditional methods.
AI benchmarks are increasingly outdated as models optimize for tests rather than true intelligence. New evaluation methods like LiveCodeBench Pro and Xbench aim to provide more meaningful measures of AI abilities.
Chinese venture capital firm Hongshan Capital Global has launched Xbench, a constantly evolving AI benchmark evaluating models on both academic tests and real-world tasks, with ChatGPT o3 leading the rankings.
Apple's critique of AI reasoning abilities is challenged by Anthropic, who argue that evaluation flaws, not model limitations, explain perceived failures in AI reasoning tasks.
New research from Apple reveals why Large Language Models tend to overthink simple puzzles but struggle and give up on complex ones, highlighting challenges in AI reasoning capabilities.
ALPHAONE introduces a universal framework to optimize AI reasoning by controlling transitions between slow and fast thinking, significantly improving accuracy and reducing computational effort across various benchmarks.
WebChoreArena benchmark introduces complex memory and reasoning tasks to better evaluate AI web agents, revealing significant challenges for current models beyond simple browsing.
NVIDIA introduces ProRL, a novel reinforcement learning method that extends training duration to unlock new reasoning capabilities in AI models, achieving superior performance across multiple reasoning benchmarks.
Microsoft's Phi-4-reasoning demonstrates that high-quality, curated data can enable smaller AI models to perform advanced reasoning tasks as effectively as much larger models, challenging the notion that bigger models are always better.
Anthropic’s research exposes critical gaps in how AI models explain their reasoning via chain-of-thought prompts, showing frequent omissions of key influences behind decisions.
Dream 7B introduces a diffusion-based reasoning approach that enhances AI's ability to reason, plan, and generate coherent text, outperforming traditional autoregressive models.
Tsinghua University researchers developed the Absolute Zero paradigm to train large language models without external data, using a self-evolving code executor system to enhance AI reasoning and learning.
Xiaomi's MiMo-7B is a compact language model that surpasses larger models in math and code reasoning through advanced pre-training and reinforcement learning strategies.
Gemini Robotics combines cutting-edge AI reasoning with physical world interaction, enabling robots to perform complex tasks with precision and adaptability.
OpenAI’s new o3 and o4-mini models introduce powerful multimodal reasoning and tool integration capabilities, enhancing AI’s accuracy and versatility across complex tasks involving text, images, and code.