Sign in to view Eric’s full profile
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
Sign in to view Eric’s full profile
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New York, New York, United States
Sign in to view Eric’s full profile
Eric can introduce you to 10+ people at Meta
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
7K followers
500+ connections
Sign in to view Eric’s full profile
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
View mutual connections with Eric
Eric can introduce you to 10+ people at Meta
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
View mutual connections with Eric
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
Sign in to view Eric’s full profile
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
About
Semantic…
Welcome back
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
Activity
7K followers
-
Eric Smith reacted on thisEric Smith reacted on thisLast week I had my last day at FAIR after six years. I had a fantastic time at FAIR and did the best work of my research career there. I'm very grateful to all my friends and colleagues that made FAIR such a wonderful place to work and brought great new ideas every day. This week I'm starting work as a Member of the Technical Staff with the founding team of AMI - Advanced Machine Intelligence, reporting to Saining Xie in New York. The team is ambitious and very talented. I am looking forward to see what we are able to do together!
-
Eric Smith reacted on thisEric Smith reacted on thisI was impacted by the Meta Superintelligence Labs layoffs yesterday. I’m an AI researcher passionate about training large language models to be both more capable and safer. Over the past few years, I’ve worked on projects advancing LLM safety, alignment, and knowledge, including: • Research and development of Llama Guard, Meta’s multimodal LLM safety guardrail • Continued pre-training, post-training, and preference alignment of Llama for production use cases (at Broadcom) I am proud of the progress we made toward safer AI systems and thankful for the brilliant researchers I had the chance to collaborate with. I am now seeking new opportunities in LLM training, alignment, or safety research. If you know of teams working on these challenges, I’d love to connect.
-
Eric Smith reacted on thisEric Smith reacted on thisNew Chapter at Cranium! I am very excited to announce that I am starting a new job as Senior Product Manager at Cranium, under the lead of Maxwell Monley. I cannot be happier of how my long job search ended up: I am looking forward to drive Product development for AI governance and security - what I believe are the most important aspects of AI today. I want to take this opportunity also to thank Anna Maria Brunnhofer-Pedemonteand Justin Bercich, PhD for letting me work with them throughout the summer at Impact AI: I wish this startup all the success, and I hope AIProdOps will become a standard soon! Leaving Meta was not easy. Like many, I’ve experienced how layoffs can shake one’s sense of stability and identity. It’s a difficult, often dehumanizing process — but I have been extremely privileged to have such incredible friends and colleagues that have been supporting me, helping networking, and just being there for me when I felt down and demoralized. There has been so many people that have been there for me I cannot list you all, but I want to call out Joshua Saxe and Chloé Bakalar for your support - and for the introductions that landed me this job! Now, it’s time for a new adventure — building products that keep AI safe, human-centered, and responsibly built.
-
Eric Smith reacted on thisSusan R. Zuber (formerly Susan Rochelle Epstein)
Susan R. Zuber (formerly Susan Rochelle Epstein)
1yEric Smith reacted on thisIt’s been an incredible journey playing a leading role at Meta in AI governance and ensuring AI works for all. After nearly four years, I’ve decided it’s time to move on to my next career chapter. Having studied AI at Stanford long before it was hot, it's been exciting to collaborate with executives, cutting-edge AI researchers, and product leaders on the policies, design, evaluation, and other go-to-market processes of many of our end to end AI systems. I'm deeply appreciative of Roy Austin entrusting me to lead Meta's civil rights work in AI these past few years. At this time, I am drawn to pursue other external opportunities. I’ll remain at Meta to help with the transition through March, and then I will move on to my next professional adventure. My upcoming endeavors will continue to center on meaningful impact including my board service and teaching at Chicago Law. Working alongside visionary leaders as we navigate the impacts of powerful and rapidly-evolving technology has been intellectually challenging and simultaneously humbling and inspiring. To my teammates, Roy, Cynthia, Ruchika, Kayla, Bobby, Manar, Julie, Leah, Rashedah, Jade, Tashiana, Liz, and especially Alessandro: our strategic discussions and shared commitment to ethical innovation and products that lift all people have been truly exceptional. I'm grateful to have worked with such talented people -- product managers, attorneys, engineers, executives, researchers, data scientists, policymakers, and so many more. My gratitude extends to the remarkable colleagues inside and outside Meta who have been part of this journey (links to colleagues are in my first comment below). I look forward to our paths intersecting again. -
Eric Smith reacted on thisEric Smith reacted on thisA New Day! A place of deep gratitude 28 years ago today, I arrived in the USA, the land of opportunities, as a young woman full of hopes and aspirations, seeking higher education and deeper scientific grounding. I have been and still am very blessed, thank God! Tomorrow marks a new chapter in my professional career. I am thrilled to announce that tomorrow I officially start as the new Director of the Language Technologies Institute (LTI) in the School of Computer Science (SCS) at Carnegie Mellon University (CMU). For those who don't know SCS at CMU, it is ranked the #1 in AI (my speciality) in the USA. The honor of being selected for the role by CMU and LTI is not lost on me. I truly appreciate their vote of confidence. They bought into my vision of critically building the next generation of talent and technologies framed in what I coin "Responsible Thinking" (more to come in a press release on Wednesday by CMU). So yes I am moving back to academia after almost 6 years of industrial experience (Thank you Meta/Facebook, namely Necip Fazil Ayan, Thank you Amazon AWS AI, namely Hassan Sawaf) where I learned so much, I grew to appreciate the impact our technologies have on people and societies, having had the privilege to work alongside some of the smartest and most talented technologists in the world. But I also realized that we have a significant gap in our training of the needed talent to address the challenges posed by these new transformative technologies. Hence the move back to the source. It was my calling. I could not have wished for a better platform than CMU, a trailblazer in the world of academia, AI and computer science. I would love to have the opportunity to harness our collective knowledge to empower the next generation of AI practitioners unlocking even more scientific creativity but with proactive responsible thinking. So YES I am grateful! I recognize that I stand on the shoulders of giants: from my PhD advisor, mentor, and friend Philip Resnik who always believed in me, my mentors Dan Jurafsky, Kathleen McKeown, Julia Hirschberg, Martha Palmer, Pascale Fung Owen Rambow, Bonnie Dorr, and a special thank you to my brother from another mother Nizar Habash, you are my go to people who have always had my back. Ever so grateful. Of course, my students and collaborators over the years, I wouldn't have this privilege without you. You helped me shape my research and accomplish what I have. It takes a village! Finally, I am grateful to my husband Kamel Ayari, my mother Hoda Khalil and my brother Hatim Diab for their support in countless ways throughout, I don't and never take it for granted. May God bless you for me! Last but not least, this is for my late father, Dr. Talat Diab, who never wavered in supporting me, his little "cowgirl", May God bless his soul! So yes a young girl can dream! My motto: She believed she could, so she did (with the help of many many people)! ما شاء الله والحمدلله، اللهم إجعل عملي لوجه الله
Experience & Education
-
Meta
******** ********** ********** **
-
**** *****
******** ******** ** ******* ******** *********** **** **********
-
******* **** *******
******
-
********* **********
****** ** ********** ******* ******* undefined
-
-
******** ********** ** *** **** ** *** ****
******** ** **** ****** *******
-
View Eric’s full experience
See their title, tenure and more.
Welcome back
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
or
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
Licenses & Certifications
-
Neural Networks for Machine Learning
Coursera Course Certificates
Volunteer Experience
Recommendations received
1 person has recommended Eric
Join now to viewView Eric’s full profile
-
See who you know in common
-
Get introduced
-
Contact Eric directly
Other similar profiles
Explore more posts
-
Evan Peterson
Thomson Reuters • 629 followers
I've recently taken a 180 w.r.t. agentic coding. It happened when my coworker put up a PR authored by Claude Opus 4.5 which implemented a complicated LLM eval workflow in one shot, and which was ~95% correct. When I leverage coding agents, code review becomes the new time-to-ship bottleneck, instead of code writing. A developer can become an architect, coordinating an army of agentic devs. And its not limited to opening a first draft PR. GitHub MCP supercharges addressing PR comments. Datadog MCP supercharges observability tasks. There are other examples. But a machine can't be held accountable for the decisions it makes. We are accountable for the code LLMs create for us. So its our responsibility to fix, iterate on, and verify what they produce, until we stand by their diffs. This is just the next step for me. Like I said, code review is the *new* bottleneck. Still thinking about how to lift that bottleneck e.g. empirical validation over opaque code via trusted testing and red teaming, etc.
27
2 Comments -
Sai Ashish Somayajula
Oracle • 4K followers
Excited to share that the GRAIL-V (Grounded Retrieval and Agentic Intelligence for Vision-Language) Workshop at CVPR 2026 is now accepting submissions. This workshop focuses on advancing grounded, retrieval-augmented, and agentic intelligence across multimodal systems — bringing together research spanning Computer Vision, NLP, and Information Retrieval. 🔥 Keynotes from Kristen Grauman (The University of Texas at Austin) Mohit Bansal (The University of North Carolina at Chapel Hill) Dan Roth (University of Pennsylvania) Scott Wen-tau Yih (Meta) 🎤 Industry Panel Moderated by Sujith Ravi (VP, Oracle AI) 📢 Call for Papers Now Open If you're working on multimodal reasoning, retrieval-integrated systems, vision-language agents, or scalable AI systems, please consider submitting your work here. 📅 Submission Deadline: March 5, 2026 🔗 Workshop Site: https://lnkd.in/g9_r625X 🔗 Submission: https://lnkd.in/gjYjqPpG Proud to see colleagues at Oracle helping push forward research at the intersection of multimodal intelligence and production-scale AI. Amit Agarwal, Hitesh Laxmichand Patel, Jyotika Singh, Vivek Gupta. #CVPR2026 #AI #ComputerVision #Multimodal #Agents #Research
8
-
Nishaanth Reddy
Apple • 2K followers
The AI industry is starting to look a lot like a food chain. Last month, Anthropic accused Moonshot AI of using hundreds of fraudulent accounts to distill Claude's capabilities to train their Kimi models. Moonshot released Kimi K2.5 as an open-weight model with a license requiring attribution for any product exceeding $20M/month in revenue. Yesterday, a developer discovered that Cursor's brand-new "Composer 2" appears to be Kimi K2.5 with reinforcement learning — the internal model ID (kimi-k2p5-rl-0317-s515-fast) leaked through the API. Cursor's parent company Anysphere reportedly does over $2B ARR. There is apparently no Kimi K2.5 branding anywhere in the product. The entire modern AI stack seems to be built on layers of borrowed intelligence. Pre-training on the internet's corpus, distillation from stronger models, RL on top of open weights and fine-tuning on synthetic data generated by competitors. At every layer, someone is building on someone else's work. These are truly interesting times. PS: Some AI was used to rewrite this.
23
-
Will Wolf
Cohere • 2K followers
🧩 Looking for a new game to try? Skeptical of AI? See if you can beat my RL agents that taught themselves Monopoly Deal! 🎉 https://monopolydeal.ai Into the technical details? Check out the paper and blog post below, detailing the platform, policy gradient methods, and "tricks" (multiple epochs per update, shared policy–value parameters, entropy regularization and decay, gradient clipping, and more...) we had to use to get this to work! 📜 https://lnkd.in/eacx8eV5 💻 https://lnkd.in/ec2YvJK3
8
-
Jessica H.
Protegrity • 4K followers
OSS Bench for Agentic Coding.. in steps Cline-Bench: "A task that frontier LLMs struggle to complete are ideal candidates for cline-bench. If you find yourself having to intervene, or write code manually because cline was not able to complete the job - your task captures the failure boundary of state-of-the-art models and is precisely what cline-bench seeks to formalize." https://lnkd.in/gHcGYxiZ
6
-
Alborz Geramifard
LinkedIn • 9K followers
🤔 Have you wondered which model size to fine-tune for your Agentic RL effort? Our recent paper (https://lnkd.in/gfFE8aUS) suggests that with the right reward shaping, smaller models is the smarter choice. RL fine-tuned Qwen3-8B matched its 32B counterpart while using 3.5× less compute, 1.5× less memory, and lower variance on TravelPlanner. It reached 56.9% final-pass with only 180 training queries — 2.7× stronger than GPT-5. These models also generalized well across out-of-domain tasks including Multi-IF, NaturalPlan, and Tau-Bench.
217
1 Comment -
Shaun Bedingfield
4K followers
This is another good post. Not complete as it doesn't mention RLVR (reinforcement learning with verifiable results) but it basically highlights that trained LLMs do arithmetic in general by statistical means (some offload to rag calculators which comes with it's own set of problems). To be fair, arithmetic should not be conflated with mathematics. The ability to do searches of literature to aid in proofs may be much more useful in the future than such a simple operation.
3
1 Comment
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content