Kangheng Liu

kanghengliu's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

updated a collection 25 days ago

CORE-bench v1.1

Collection

Benchmark for AI agents on scientific reproducibility — mainline (39) and OOD (19) splits derived from Code Ocean capsules. • 2 items • Updated 25 days ago

updated 2 datasets 25 days ago

agent-evals/core-bench-v1.1-ood

Viewer • Updated 25 days ago • 19 • 95

agent-evals/core-bench-v1.1-mainline

Viewer • Updated 25 days ago • 39 • 144

published 2 datasets 27 days ago

agent-evals/core-bench-v1.1-ood

Viewer • Updated 25 days ago • 19 • 95

agent-evals/core-bench-v1.1-mainline

Viewer • Updated 25 days ago • 39 • 144

upvoted a paper 3 months ago

Towards a Science of AI Agent Reliability

Paper • 2602.16666 • Published Feb 18 • 16

updated a dataset 4 months ago

agent-evals/hal_traces

Updated Feb 1 • 1.13k • 5

updated a dataset 5 months ago

agent-evals/hal_traces

Updated Feb 1 • 1.13k • 5

updated a dataset 6 months ago

agent-evals/hal_traces

Updated Feb 1 • 1.13k • 5