#mathematical-reasoning

[ follow ]
#multimodal-ai #vision-language-models #model-efficiency #ai-agents #gpt-52-pro #model-benchmarking #epoch-ai
Software development
fromTechzine Global
6 days ago

Microsoft introduces open-source multimodal Phi-4 reasoning model

Microsoft's Phi-4-reasoning-vision-15B combines vision and reasoning capabilities using mid-fusion architecture, outperforming larger models on mathematical and scientific benchmarks while maintaining efficiency through selective multimodal layer processing.
fromComputerworld
1 month ago

OpenAI's GPT is getting better at mathematics

OpenAI's GPT-5.2 Pro does better at solving sophisticated math problems than older versions of the company's top large language model, according to a new study by Epoch AI, a non-profit research institute.
Artificial intelligence
Artificial intelligence
fromTechCrunch
1 month ago

AI models are starting to crack high-level math problems | TechCrunch

Advanced LLMs like GPT-5.2 can solve open mathematical problems and produce novel, verifiable proofs that extend mathematical research.
Artificial intelligence
fromInfoQ
2 months ago

Intel DeepMath Introduces a Smart Architecture to Make LLMs Better at Math

DeepMath uses a Qwen3-4B Thinking agent that emits small Python executors for intermediate math steps, improving accuracy and significantly reducing output length.
fromNature
3 months ago

DeepSeek's self-correcting AI model aces tough maths proofs

The model, DeepSeekMath-V2, scored 118 out of 120 points on questions from the 2024 William Lowell Putnam Mathematical Competition, beating the top human score of 90. The model also performed at the level of gold-medal winners in the International Mathematical Olympiad (IMO) 2025 and the 2024 China Mathematical Olympiad. The results are described in a preprint posted on arXiv on 27 November.
Artificial intelligence
Artificial intelligence
fromNature
3 months ago

DeepSeek's self-correcting AI model aces tough maths proofs

DeepSeekMath-V2 scored 118/120 on the 2024 Putnam, surpassing top humans and using self-verifiable reasoning to detect and correct its own errors.
Artificial intelligence
fromArs Technica
3 months ago

DeepMind's latest: An AI for handling mathematical proofs

AlphaProof achieved International Mathematical Olympiad silver-level performance and nearly gold on the Putnam, demonstrating substantial advances in automated mathematical reasoning.
fromstupidDOPE | Est. 2008
7 months ago

Google's Gemini 2.5 AI Model Launches with Major Upgrades for Ultra Users | stupidDOPE | Est. 2008

Gemini 2.5 stands out from other AI offerings thanks to its multi-agent structure. This design enables the model to simulate multiple AI agents that work together to analyze, test, and refine solutions to a task.
Artificial intelligence
[ Load more ]