LEAF: Growing Trees Without Branching for Speech-Aware Large Language Model Post-Training

In three linesLEAF is a retrospective tree-based RL method for speech-aware LLM post-training that improves credit assignment by grouping responses by shared prefixes and assigning span-level advantages. It outperforms GRPO on speech question-answering and speech translation benchmarks without online branching or additional decoding.

Summary generated by Claude — human-verified