AllSpark Research · 10 records

dateidtitleauthorsvenue↑
2026-10-142610.14709Vela: User Behavior as Supervision Sparks LLM IntelligenceYang Xiao et al.arXiv85
2026-09-302609.35954ROSS: Relearning from Self-Generated Rollouts through Selective SupervisionZhiwei Zhang et al.arXiv34
2026-09-292609.33665CompoWorld: Compositional Environment Scaling for General AgentsXiao-Wen Yang et al.arXiv40
2026-09-292609.33772Skill2Env: Capability-Oriented Environment Synthesis from Skills for General AgentsWeiyi Xu et al.arXiv36
2026-09-292609.33642Learning to Learn from Context: Synthetic Training from Perturbed Public DocumentsHaoyi Wu et al.arXiv31
2026-09-242609.26355PACT: From Credit Assignment to Critic AlignmentJiayan Fu et al.arXiv28
2026-09-082609.02998Verify Before You Distill: Prompt-Level Teacher Gating for On-Policy DistillationZhiwei Zhang et al.arXiv16
2026-09-072609.04304Iris: Climbing to the Search FrontierZiyuan Liu et al.arXiv50
2026-08-282608.26530PILOT in the Loop: Live Self-Improvement for Long-Horizon AgentsYang Xiao et al.arXiv36
2026-08-272608.24987D^3-MOPD: Adaptive Dynamic Domain ScheDuling for Efficient Multi-Teacher DistillationZechen Sun et al.arXiv28