Tag

#reasoning

12 articles

Create a Reasoning-Focused LLM: A Practical Guide to Streaming, Curating, and Fine-Tuning the SupraLabs Reasoning Corpus

Learn how reasoning-focused language models are built to think through problems step-by-step, making AI more useful and trustworthy. This beginner-friendly guide explains the process using real-world examples.

Aug 1344

SpaceXAI Releases Grok 4.6: A 500K-Context Frontier Model Tuned for Long-Running Agents, Coding, and Knowledge Work

This article explains the technical advancements in SpaceXAI's Grok 4.6, focusing on its 500K token context window, xhigh reasoning level, and post-training optimization techniques.

Aug 1252

Cogent AI Team Releases VR-1: A Frontier Cyber Reasoning Model That Composes and Verifies Enterprise Attack Paths

This article explains VR-1, a new AI model designed for cybersecurity reasoning that composes and verifies enterprise attack paths. It covers the model's architecture, functionality, and significance in modern cybersecurity.

Aug 243

Sina's open model VibeThinker-3B aims to show reasoning compresses well but factual knowledge doesn't

A new AI model called VibeThinker-3B shows that logical reasoning can be made small and efficient, while factual knowledge still requires larger models. This discovery could make AI more accessible and powerful.

Jun 2770

VibeThinker-3B: A 3B Dense Reasoning Model Built on Qwen2.5-Coder-3B With the Spectrum-to-Signal Post-Training Pipeline

Learn how to load, test, and evaluate the VibeThinker-3B reasoning model using Hugging Face transformers and Python.

Jun 1985

OpenAI researchers explain why math is the road to AGI

OpenAI researchers Sebastian Bubeck and Ernest Ryu explain how mathematical reasoning is becoming a critical indicator in the pursuit of artificial general intelligence (AGI).

Apr 2969

Alibaba's Qwen team makes AI models think deeper with new algorithm

Alibaba's Qwen team introduces a novel algorithm that enhances AI reasoning by weighting each step based on its influence on subsequent decisions, effectively doubling the length of thought processes in AI models.

Apr 4143

GPT reasoning models have "line of sight" to AGI, says OpenAI's Greg Brockman

This explainer explores the concept of General Artificial Intelligence (AGI) and how OpenAI's Greg Brockman believes GPT reasoning models are on a clear path toward achieving it, using the term 'line of sight' to describe this trajectory.

Apr 2119

Anthropic confirms leaked model marks a "step change" in reasoning after data breach reveals its existence

This article explains how a security leak revealed Anthropic's advanced AI model, Claude 3.5 Sonnet, and what this means for the future of AI reasoning capabilities.

Mar 26129

Luma Labs Launches Uni-1: The Autoregressive Transformer Model that Reasons through Intentions Before Generating Images

Explains how Luma Labs' Uni-1 model introduces a reasoning phase before image generation, addressing the 'intent gap' that affects current diffusion models.

Mar 23107

AI models can barely control their own reasoning, and OpenAI says that's a good sign

OpenAI introduces 'CoT controllability' as a new safety measure, finding that AI models struggle to control their own reasoning—yet the company sees this as a positive development for AI safety.

Mar 6173

OpenAI launches GPT-5.4 Thinking and Pro combining coding, reasoning, and computer use in one model

OpenAI launches GPT-5.4 Thinking and Pro, combining coding, reasoning, and computer operation into a single AI model for the first time.

Mar 5135