Twitxy Saved

rLLM

Verified account This account is verified. Learn more

@rllm_project

Agentic RL training over any harness and environments | @BerkeleySky

Github: github.com/rllm-org/rllm
rllm-project.com Joined October 2025
14 Following 1.1K Followers 92 Posts

Latest posts 5

rLLM Verified account @rllm_project
1/5 New rLLM Blog Post: Continual learning for agents 🧵

Can an AI agent keep learning after deployment—from real production traffic?

Each interaction yields one trajectory. Unlike build-time RL, you typically can't sample multiple rollouts for the same prompt, so GRPO's group-relative advantage no longer applies.

Our result shows a +16.2 points improvement on real SWE tasks.

Full post: rllm-project.com/post.html?post…
2 replies 5 reposts 19 likes 4.3K views View on X
rLLM Verified account @rllm_project
🇧🇷 Olá, Rio de Janeiro!
A few members of our rLLM team (@StarLi27496427, @kylepmont, @jaraxiong) are joining #ICLR2026 from April 23–27.

Whether you're a user of rLLM 🛠️, curious about our recent Hive agent swarm 🐝, or simply want to chat about agentic RL & LLM post-training 🤖 — feel free to DM us and meet the team!

Hope everyone has an amazing week in Rio🏖️, and stay tuned for some exciting updates ahead!
0 replies 3 reposts 11 likes 655 views View on X