Search papers, labs, and topics across Lattice.
3
0
8
Forget prompting monolithic models – ImageEdit-R1 uses reinforcement learning to orchestrate a team of specialized agents, outperforming even closed-source diffusion models on complex image editing tasks.
Forget SFT: this RL-only method teaches LLMs to use tools by showing them how, then gradually removing the training wheels for zero-shot mastery.
Get 50% shorter LLM responses without sacrificing accuracy using a new RL method that dynamically balances task reward and length constraints.
