Lincan Li

2 articles on SOTA Papers

Computer Science

Agent Swarm Cuts Complex Task Latency Up To 4.5×

Kimi K2.5 couples joint text-vision training with dynamically scheduled parallel subagents, improving agentic search scores while reducing time to target quality.

cs.AIcs.LGcs.CL
Aug 13, 20264 min2602.02276
Computer Science

Kimi K3 Brings Open Weights Closer to Frontier Agents

A 2.8T-parameter MoE combines Delta Attention, 16-of-896 expert routing, and agentic reinforcement learning to approach leading proprietary systems at lower task cost.

cs.LGcs.CL
Aug 10, 20264 min2607.24653