7 papers
On Surjectivity of Neural Networks: Can you elicit any behavior from your model?
Haozhe Jiang, Nika Haghtalab
Given a trained neural network, can any specified output be generated by some input? Equivalently, does the network correspond to a function that is surjective? In generative model…
Think When Needed: Adaptive Reasoning-Driven Multimodal Embeddings with a Dual-LoRA Architecture
Longxiang Zhang, Weilong Dai, Guanghao Zhang +2
Multimodal large language models (MLLMs) have emerged as a powerful backbone for multimodal embeddings. Recent methods introduce chain-of-thought (CoT) reasoning into the embedding…
Muninn: Your Trajectory Diffusion Model But Faster
Gokul Puthumanaillam, Hao Jiang, Ruben Hernandez +4
Diffusion-based trajectory planners can synthesize rich, multimodal robot motions, but their iterative denoising makes online planning and control prohibitively slow. Existing acce…
The Extrapolation Cliff in On-Policy Distillation of Near-Deterministic Structured Outputs
Xin Li, Hao Jiang, Annan Wang +2
On-policy distillation (OPD) is widely used for LLM post-training. When pushed with a reward-extrapolation coefficient lambda > 1, the student can lift past the teacher in domain,…
Towards Customized Multimodal Role-Play
Chao Tang, Jianzong Wu, Qingyu Shi +5
Unified multimodal understanding and generation models enable richer human-AI interaction. Yet jointly customizing a character's persona, dialogue style, and visual identity while…
LFPO: Likelihood-Free Policy Optimization for Masked Diffusion Models
Chenxing Wei, Jiazhen Kang, Hong Wang +8
Reinforcement Learning with Verifiable Rewards (RLVR) has achieved remarkable success in improving autoregressive models, especially in domains requiring correctness like mathemati…