4 accepted papers
Masked diffusion models (MDMs) are a promising alternative to autoregressive models (ARMs), but they suffer from **inherently** much higher training variance. High variance leads to noisier gradient estimates and unstable optimization, so even equally strong pretrained MDMs and ARMs that are competi…
Reward models (RMs) are crucial in reinforcement learning from human feedback (RLHF) to align large language models (LLMs) with human values. However, RM training data is commonly recognized as low-quality, always containing preference conflicts and inductive biases, such as response length or speak…
Reinforcement learning with verifiable rewards (RLVR) has become the mainstream technique for training LLM agents. However, RLVR highly depends on well-crafted task queries and corresponding ground-truth answers to provide accurate rewards, which requires significant human effort and hinders the sca…
Modern Transformers predominantly adopt the Pre-Norm paradigm for its optimization stability, foregoing the superior potential of the unstable Post-Norm architecture. Prior attempts to combine their strengths typically lead to a stability-performance trade-off. We attribute this phenomenon to a stru…