M2N2 is a model merging technique that creates powerful multi-skilled agents without the high cost and data needs of retraining.

Memp takes inspiration from human cognition to give LLM agents "procedural memory" that can adapt to new tasks and environments.

The open source framework provides the data and training recipe for building powerful computer-use agents that challenge proprietary systems.

Researchers from Inclusion AI and Ant Group proposed a new LLM leaderboard that takes its data from real, in-production apps.

Chain-of-Thought isn't a plug-and-play solution. For developers, this research offers a blueprint for LLM testing and strategic fine-tuning.

Morris found it could also reproduce verbatim passages from copyrighted works, including three out of six book excerpts he tried.

For enterprise teams and commercial developers, this means the model can be embedded in products or fine-tuned.

For now, the changes should help placate users who felt frustrated by the sudden shift to GPT-5 and deprecation of OpenAI's older LLMs.

The pressure is on for OpenAI to prove that GPT-5 isn’t just an incremental update, but a true step forward.

It also failed on a simple algebra arithmetic problem that elementary schoolers could probably nail, 5.9 = x + 5.11.