Jekyll2026-09-02T16:40:48+00:00https://alexanderjiang.com/feed.xmlAlexander JiangAlexander Jiang's academic websiteAlexander Jiangzander.jiang@gmail.comIsoExec: Unified Execution to Eliminate Trainer-Inference Mismatch in SkyRL2026-08-21T00:00:00+00:002026-08-21T00:00:00+00:00https://alexanderjiang.com/posts/2026/08/isoexecIsoExec is a unified execution framework that eliminates numerical discrepancies between training and inference engines in reinforcement learning workloads. Using an execution contract and batch-invariant kernels, it achieves bitwise consistency across model parallelism strategies.

Read the full post on the vLLM blog

]]>Alexander Jiangzander.jiang@gmail.comFlashInfer-Bench: Building the Virtuous Cycle for AI-driven LLM Systems2025-10-21T00:00:00+00:002025-10-21T00:00:00+00:00https://alexanderjiang.com/posts/2025/10/flashinfer-benchIntroducing FlashInfer-Bench, a benchmark and infrastructure that lets AI systems optimize themselves: standardized GPU workload descriptions via FlashInfer Trace, realistic benchmarks from production LLM deployments, and a seamless path for deploying AI-generated kernels into live serving systems.

Read the full post on flashinfer.ai

]]>Alexander Jiangzander.jiang@gmail.com