Europe’s ability to remain competitive will depend on how effectively it advances critical technologies ranging from automation and programming to AI, connectivity, clean energy, and biosciences, since these domains are shaping future growth and resilience. Microblog @antgrasso
RESEARCH
-

Comprehensive Guide to LLM Post-Training: SFT, RLHF, and RL Algorithms
By
–
An excellent technical guide on LLM post-raining covering SFT(supervised finetuning), RL rewards such as RLHF/human preferences, RLAIF/constitutional-AI, RLVR/verifiable outcomes, process-supervised and rubric rewards. Also covers common RL training algorithms from PPO, GRPO, and
-
Qwen 2.5 Release Transforms LLM Landscape Forever
By
–
Qwen 2.5 forever GOATED i don't think the LLMs landscape would be the same without that specific release
-
GPT-4 Agents Outperformed Without Weight Adjustments
By
–
The fact that it beats GPT-4 agents without touching weights… That’s a turning point.
-

Google’s Self-Organizing AI System: TUMIX
By
–
Google just built an AI that organizes itself. It’s called TUMIX, and it might be the most interesting paper Google has published this year. Instead of training a bigger model, the team built a system where multiple AIs work together at test time. Each agent uses different
-
AI Objectives and Human Bias Mimicry
By
–
L'objectif d'une IA est de mimer ce que fait un humain. Les biais en font parti…
-
AI as Industrial Revolution: Comparing Past Technological Paradigm Shifts
By
–
If you think the future is like the past, then it all comes down to your analogy: 1880s railroads? 2000 dot-com? 1910s factory electrification? 1980s software? It is also possible to argue that AI represents an industrialization-style break with all past trends. We don't know.
-
AI Future Scenarios: Bubble, Plateau, or Transformative Breakthrough?
By
–
I don't think we have a good grasp on what the future of AI will be, so it would be reasonable to start thinking in scenarios: what if there is a bubble or a plateau in AI development? (Not the same thing) Conversely, what if the insiders saying transformative AI soon are right?
-
RGBD Post-Processing Limitations for Robot Hand Motion Training
By
–
Ah! Small world! I appreciate your suggestions on Justin’s talk! For your data, I worry that current methods for post- processing RGBD aren’t yet accurate enough to recover the detailed hand and object motions needed for robot training
-

Geoffrey Hinton Confirmed as AI Legend
By
–
i always suspected geoffrey hinton was an absolute legend
