University of Science and Technology of China · 20 records

dateidtitleauthorsvenue↑
2026-11-192611.23751Thinking Fast, Anywhere: One System-1 for the Decision Points of Any Agent WorkflowRui Li et al.arXiv14
2026-11-092611.07624Beneath the Surface: What We Overlook When We Talk About Agent EvolutionXiaobo Wang et al.arXiv25
2026-11-04W4413120297An open structure-guided RNA foundation model for generalizable predictionHeqin Zhu et al.Nature
2026-10-31W7224787484Chemoreceptor TRPγ coordinates locomotor activity via regulating intracellular Ca2+ homeostasis in DrosophilaPeiyuan Pang et al.Nature Communications
2026-10-31W7225005953AnnelidaDB: a comprehensive platform of taxonomic and multi-omic resources across annelidsYunlong Li et al.Nucleic Acids Research
2026-10-31W4415421078Chemoreceptor TRPγ coordinates locomotor activity via regulating intracellular Ca2+ homeostasis in DrosophilaPeiyuan Pang et al.Nature Communications
2026-10-192610.20556Speculative Sampling in Interactive Video GenerationYubo Huang et al.arXiv55
2026-10-192610.22234OS-Helix: Unifying GUI Interaction and Code Execution in Native Computer-Using AgentsBowen Yang et al.arXiv12
2026-10-152610.20516VASPO: Self-Distillation with Privileged Value EstimationFei Wu et al.arXiv5
2026-10-142610.15166ExeWorldBench: Benchmarking Code-Based Executable 3D World ModelingXinge Peng et al.arXiv62
2026-10-022609.38923GraphForge: Training Working Agents with Graph-Anchored Workspace SynthesisQisheng Su et al.arXiv58
2026-10-022610.00559PhysVista: Benchmarking Physical Intelligence in VLMs via a Perception-Reasoning-Assessment LoopXinge Peng et al.arXiv34
2026-09-282609.04355VLA-Precision: Asymmetric Co-Bootstrapping for Efficient Real-World Online RL of Vision-Language-Action ModelsChenyu Su et al.arXiv7
2026-09-182609.12397UFO: Chain-of-Evaluation for Omni-Condition Alignment in Multi-Modal Image GenerationDanning Zhang et al.arXiv18
2026-08-212608.18580FACET: Preserving Source Intent and Executable State in Terminal Task SynthesisKou Shi et al.arXiv34
2026-06-192606.20545Current World Models Lack a Persistent State CoreJinpeng Lu et al.arXiv18
2026-06-052605.30159Meta-Cognitive Memory Policy Optimization for Long-Horizon LLM AgentsZiyan Liu et al.arXiv7
2026-05-292605.27995AsyncTool: Evaluating the Asynchronous Function Calling Capability under Multi-Task ScenariosKou Shi et al.arXiv10
2026-05-222605.21850ACC: Compiling Agent Trajectories for Long-Context TrainingQisheng Su et al.arXiv60
2026-05-152605.14876Unlocking Complex Visual Generation via Closed-Loop Verified ReasoningHanbo Cheng et al.arXiv2