Jiachen Li

Continual pretraining on AXIS boosts robot manipulation success rates by 5.8% over previous methods and outperforms existing datasets by over 37%.

StratMamba achieves a 5% faster navigation speed and 0.915 path efficiency, setting a new benchmark for obstacle avoidance in robotic navigation.

AnyBody enables humanoid robots to seamlessly control movements from any chosen subset of body keypoints, revolutionizing how we approach humanoid locomotion and manipulation.

Rule-grounded reasoning can cut average distance errors in driving VLAs by nearly half, fundamentally enhancing their decision-making transparency and reliability.

MLLM-empowered re-ranking can dramatically enhance person re-identification accuracy in unseen domains, overcoming traditional encoder limitations.

EOM-CCSD achieves unprecedented accuracy in band gap predictions, outperforming traditional methods with a mean absolute error of just 0.27 eV.

Existing corpus poisoning attacks falter in realistic RAG systems, but the new CRCP framework ensures adversarial effectiveness by aligning chunking and reranking processes.

Embodied navigation agents, already struggling, fall apart when faced with the kinds of messy, real-world sensor and instruction corruptions that NavTrust now exposes.

$GW$ self-energies have a discontinuity at integer particle numbers, explaining why they can give accurate quasiparticle energies despite large delocalization errors in RPA total energies.

LLMs can be used to prune irrelevant information *before* planning, enabling efficient long-horizon multi-robot task planning that outperforms both pure LLM and hybrid LLM-PDDL approaches.

A 1000x larger video reasoning dataset reveals early signs of emergent generalization, offering a new foundation for training and evaluating spatiotemporal AI.

Unlock human-like spatial reasoning in VLMs with VLM-3R, which reconstructs 3D understanding from monocular video using instruction tuning, bypassing the need for external depth sensors.