Search papers, labs, and topics across Lattice.
6
4
7
12
FunReason-MT is presented, a novel data synthesis framework for real-world multi-turn tool use that resolves the complexity barrier in multi-turn FC data by employing 1) Environment-API Graph Interactions to gather varied high-quality trajectories, 2) Advanced Tool-Query Synthesis to simplify hard query construction, and 3) Guided Iterative Chain for sophisticated CoT generation.
CrEST redefines credit assignment in RL by shifting the teacher's role from directing updates to modulating their magnitude, leading to substantial performance improvements in multi-turn agent training.
RODS synthesizes new training data on-the-fly, enabling agents to maintain high performance with 20x fewer trajectories than traditional methods.
LLMs still struggle to generate high-quality interactive HTML applications, despite their advancements in code generation, highlighting a gap that MiniAppBench aims to address.
Forget random sampling – this framework crafts targeted, multi-turn function-calling data that catapults smaller LLMs to state-of-the-art performance.
Forget synthetic data and overfitting: Environment Tuning lets LLM agents learn complex tool-use behaviors directly from the environment, slashing data needs and boosting generalization.
