In this blog, I explore how different large language models (LLMs) vary in their named entity recognition (NER) outputs and propose methods to unify these discrepancies using cosine similarity and Sentence-BERT. ws-dl.blogspot.com/2025/01/2… @WebSciDL @phonedude_mln @storymodelers

Xiaohongshu, or Rednote, the Chinese Instagram, is now #1 in the US app store after the TikTok ban. Here's the most interesting paper on how their feed algorithm works.

After 21 years of leadership, dedication, and service, we would like to extend our heartfelt congratulations to OERI Chief Operations Officer, Dr. Mike Robinson, on his well-deserved retirement. More here: oduinnovate.org/oeri-announc… #ODUInnovate

Don't do RAG Proposes cache-augmented generation (CAG) to eliminate retrieval latency and minimize retrieval errors. What is CAG? CAG aims to leverage the capabilities of long-context LLMs by preloading the LLM with all relevant docs in advance and precomputing the key-value (KV) cache. The preloaded context helps the model to provide contextually accurate answers without the need for additional retrieval during runtime. When to apply CAG? It's a useful alternative to RAG for cases where the documents/knowledge for retrieval are of limited, manageable size. My thoughts: As LLMs advance in capabilities, I suspect that what we know as RAG today could change significantly either architecturally or how it's optimized. CAG is one in a growing list of developments and new ideas that have emerged recently to address limitations like poor retrieval relevancy and latency. There could also be hybrid methods that combine preloading with selective retrieval. Don't sleep on long-context LLMs. They are here to stay.

Storytelling matters in econ too

We are about to announce the launch of a new Master's Degree in #Computational Social and Political Science at @LaStatale in Milan, starting in September 25. Please share this info and my contact to any student potentially interested! More soon!

Part of our @LaStatale #BehaveLab team at a dinner in a trattoria in Milan to celebrate the end of a long and productive year! @squazzoni @federico_fb @ElenaDeGioannis @raffaelevacca

Today was the final day of our Mission Engineering and Integration workshop. It was a unique learning experience as there are only a few places, resources, and people who teach Mission Engineering. The workshop started with the introduction of key Mission Engineering concepts, followed by hands-on learning in multidisciplinary teams. The participants concluded the workshop by applying the knowledge acquired during the last three days to generate a mission thread. We hope that the skills acquired during the workshop will make the participants Mission Engineering ambassadors in their organizations. Stay tuned for information about the next @MissionHubODU workshop!