Fudan University

Trends

• Agent post-training becomes predictable engineering

Items

1. Aug 15

   Fudan's SWE-bench Science stumps top agents

   A repository-level benchmark of 119 scientific-software tasks across 98 repositories and 20 domains on which the strongest agent, Claude Code with Opus 5 at max effort, scores below 50% pass@1, with scientific guidance helping only when well aligned.

2. May 5

   Fudan's Balanced Aggregation fixes GRPO bias

   OpenMOSS identifies sign-length coupling and length downweighting as twin aggregation biases in GRPO and fixes both with a parameter-free drop-in, improving stability and scores for Qwen2.5-Math-7B and Qwen3-1.7B across six reasoning and coding benchmarks.

3. Apr 18

   Fudan's GenericAgent

   A self-evolving agent that maximizes decision-relevant information within a sub-30K context using nine atomic tools, hierarchical on-demand memory, and verified trajectories crystallized into reusable SOPs and code, outperforming agents that consume 200K-1M tokens on task completion and web browsing.