Li Chen
San Francisco, California, United States
753 followers
500+ connections
About
RESEARCH PROFILE
* Research scientist and technical lead with a Ph.D. in Computer…
Activity
-
Escaping the Agreement Trap: Defensibility Signals for Evaluating Rule-Governed AI Excited to share our new preprint with Rosa Català In content…
Escaping the Agreement Trap: Defensibility Signals for Evaluating Rule-Governed AI Excited to share our new preprint with Rosa Català In content…
Liked by Li Chen
-
After nearly a year at Meta working on Search, I've decided to head back to Google in a few weeks. I'll share more once I'm there. My time at Meta…
After nearly a year at Meta working on Search, I've decided to head back to Google in a few weeks. I'll share more once I'm there. My time at Meta…
Liked by Li Chen
-
Not one, but five papers on relational data at #ICLR2026 🇧🇷 this week in Rio. Relational Graph Transformer: April 24 | Pavilion 3, Poster Session…
Not one, but five papers on relational data at #ICLR2026 🇧🇷 this week in Rio. Relational Graph Transformer: April 24 | Pavilion 3, Poster Session…
Liked by Li Chen
Experience
Education
-
Virginia Polytechnic Institute and State University
-
-
Machine Learning
Computer Vision
Image/Signal Processing
Statistical Modeling -
-
-
-
-
Languages
-
English
-
-
Chinese
-
More activity by Li
-
Another milestone for Worki: we’ve closed our pre-seed round! At the same time, we received strong praise from a key customer this month — that’s…
Another milestone for Worki: we’ve closed our pre-seed round! At the same time, we received strong praise from a key customer this month — that’s…
Liked by Li Chen
-
⚡ Meet Qwen3.6-35B-A3B:Now Open-Source!🚀🚀 A sparse MoE model, 35B total params, 3B active. Apache 2.0 license. 🔥 Agentic coding on par with…
⚡ Meet Qwen3.6-35B-A3B:Now Open-Source!🚀🚀 A sparse MoE model, 35B total params, 3B active. Apache 2.0 license. 🔥 Agentic coding on par with…
Liked by Li Chen
-
Today we're releasing Muse Spark, the first model from MSL. Nine months ago we rebuilt our AI stack from scratch. New infrastructure, new…
Today we're releasing Muse Spark, the first model from MSL. Nine months ago we rebuilt our AI stack from scratch. New infrastructure, new…
Liked by Li Chen
-
I've never built an app before, and I've never thought of myself as a product person, and now here we are. I'm building a dead simple app I've…
I've never built an app before, and I've never thought of myself as a product person, and now here we are. I'm building a dead simple app I've…
Liked by Li Chen
-
Squeezing more out of every GPU, introducing our new model co-hosting serving container solution. It allows multiple model instances to be served…
Squeezing more out of every GPU, introducing our new model co-hosting serving container solution. It allows multiple model instances to be served…
Liked by Li Chen
-
V-JEPA 2.1: a new JEPA model trained on video. Uses features from multiple levels in the hierarchy. Better performance than V-JEPA 2 on some dense…
V-JEPA 2.1: a new JEPA model trained on video. Uses features from multiple levels in the hierarchy. Better performance than V-JEPA 2 on some dense…
Liked by Li Chen
-
Grateful for the opportunity to conduct a recent inspection visit at the outstanding Moffitt Cancer Center . It was a pleasure connecting with such…
Grateful for the opportunity to conduct a recent inspection visit at the outstanding Moffitt Cancer Center . It was a pleasure connecting with such…
Liked by Li Chen
-
Introducing Shoply Site Chat: Think Google’s site search - but fully conversational. With just a few lines of configuration to specify which websites…
Introducing Shoply Site Chat: Think Google’s site search - but fully conversational. With just a few lines of configuration to specify which websites…
Liked by Li Chen
-
🚀 Introducing the Qwen 3.5 Medium Model Series Qwen3.5-Flash · Qwen3.5-35B-A3B · Qwen3.5-122B-A10B · Qwen3.5-27B ✨ More intelligence, less compute.…
🚀 Introducing the Qwen 3.5 Medium Model Series Qwen3.5-Flash · Qwen3.5-35B-A3B · Qwen3.5-122B-A10B · Qwen3.5-27B ✨ More intelligence, less compute.…
Liked by Li Chen
-
🚀 Qwen3.5-397B-A17B-FP8 weights are now open! It took some time to adapt the inference frameworks, but here we are: ✅ SGLang support is merged 🔄…
🚀 Qwen3.5-397B-A17B-FP8 weights are now open! It took some time to adapt the inference frameworks, but here we are: ✅ SGLang support is merged 🔄…
Liked by Li Chen
-
Excited to share that I built an agent-only trading platform in just two days! Looking forward to the future of trading with AI. Check it out at…
Excited to share that I built an agent-only trading platform in just two days! Looking forward to the future of trading with AI. Check it out at…
Liked by Li Chen
-
Our superhuman reasoning team at Google DeepMind has been hard at work over the past year to use AI, in collaboration with mathematicians, towards…
Our superhuman reasoning team at Google DeepMind has been hard at work over the past year to use AI, in collaboration with mathematicians, towards…
Liked by Li Chen
-
Exciting news to kick off the new year: I’ve recently joined LMArena as Head of Data & Applied ML, where I’ll be helping build the platform that…
Exciting news to kick off the new year: I’ve recently joined LMArena as Head of Data & Applied ML, where I’ll be helping build the platform that…
Liked by Li Chen
-
We’ve relied on OpenAI’s GPT-4.1 for most of the past year, but Grok 4.1 has genuinely blown us away. For several AI Search use cases, Grok 4.1…
We’ve relied on OpenAI’s GPT-4.1 for most of the past year, but Grok 4.1 has genuinely blown us away. For several AI Search use cases, Grok 4.1…
Liked by Li Chen
-
Meeting with SAMA(Saudi Arabian Monetary Agency). I firmly believe middle east will become another center of AI innovation.
Meeting with SAMA(Saudi Arabian Monetary Agency). I firmly believe middle east will become another center of AI innovation.
Liked by Li Chen
Other similar profiles
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content