Reasoning

18 Articles
New metric tracks where multimodal reasoning models go wrong
Tech

New metric tracks where multimodal reasoning models go wrong

(a) Example of outputs from a reasoning model and a non-reasoning model on a perception task. Red highlights indicate visual hallucination. Multimodal reasoning...

Vision-language models gain spatial reasoning skills through artificial worlds and 3D scene descriptions
Tech

Vision-language models gain spatial reasoning skills through artificial worlds and 3D scene descriptions

On the left, the simulated environment containing a cuboid placed on a plane and observed by a camera, placed directly above the object...

Multi-modal AI agent mimics human thinking for long video analysis and reasoning
Tech

Multi-modal AI agent mimics human thinking for long video analysis and reasoning

Credit: GitHub: While Artificial Intelligence (AI) technology is evolving rapidly, AI models still struggle with understanding long videos. A research team from The...

Cassie’s Reasoning For Not Leaving Diddy Allegedly Involved JAY-Z
Foreign Celebrity

Cassie’s Reasoning For Not Leaving Diddy Allegedly Involved JAY-Z

Sean “Diddy” Combs’ former assistant alleges that JAY-Z‘s relationship with Beyoncé influenced Cassie‘s decision to stay in an allegedly abusive relationship with the...

GPT-4 matches human performance on analogical reasoning tasks, study shows
Tech

GPT-4 matches human performance on analogical reasoning tasks, study shows

Results for letter-string analogies with shuffled alphabet. Credit: PNAS Nexus (2025). DOI: 10.1093/pnasnexus/pgaf135 Can large language models (LLMs) reason by analogy? Some outputs...

Reinforcement learning boosts reasoning skills in new diffusion-based language model d1
Tech

Reinforcement learning boosts reasoning skills in new diffusion-based language model d1

Log Probability Estimation in diffu-GRPO. Credit: arXiv (2025). DOI: 10.48550/arxiv.2504.12216 A team of AI researchers at the University of California, Los Angeles, working...

OpenAI beats DeepSeek on sentence-level reasoning
Tech

OpenAI beats DeepSeek on sentence-level reasoning

Credit: AI-generated image ChatGPT and other AI chatbots based on large language models are known to occasionally make things up, including scientific and...

Deep Reasoning is coming to ChatGPT free, but I think it’s still worth paying for ChatGPT Plus
Tech

Deep Reasoning is coming to ChatGPT free, but I think it’s still worth paying for ChatGPT Plus

Deep Research is coming to the free tier of ChatGPT very soon The Plus tier still offers considerable advantages over the free tier...

Gemini’s ‘most intelligent AI model’ yet is now available for free – here are 3 ways you can use its incredible reasoning capabilities
Tech

Gemini’s ‘most intelligent AI model’ yet is now available for free – here are 3 ways you can use its incredible reasoning capabilities

Google announced Gemini 2.5 last week Now you can access its reasoning model, Gemini 2.5 Pro Experimental, for free It tops Humanity’s Last...