PPL
Haiyang Shen
Papers (6)
EvoCode-Bench: Evaluating Coding Agents in Multi-Turn Iterative InteractionsSGR-Bench: Benchmarking Search Agents on State-Gated RetrievalMindLoom: Composing Thought Modes for Frontier-Level Reasoning Data SynthesisDeepWeb-Bench: A Deep Research Benchmark Demanding Massive Cross-Source Evidence and Long-Horizon DerivationTeaching AI Through Benchmark Construction: QuestBench as a Course-Based Practice for Accountable Knowledge WorkRoadmapBench: Evaluating Long-Horizon Agentic Software Development Across Version Upgrades