Unverified50% confidenceFactExact time
DataBugBench is a benchmark for evaluating large language models' debugging ability on invisible bugs in data science.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedDataBugBench specifically constructs diagnostic tests targeting Pandas, NumPy, and Scikit-Learn code.72% similarUnverifiedLong-horizon coding benchmarks (such as SWE-bench) evaluate a model's ability to handle repository-level tasks, including reading codebases, locating bugs, making cross-file modifications, and passing tests.71% similarUnverifiedData API testing should version-control request parameters, response samples, quota consumption, and assertion conditions together rather than relying on the developer's memory.70% similarUnverifiedPromptfoo can cover all three evaluation layers—large models, knowledge bases, and agents—making it suitable for testing teams whose purpose is to determine whether software can go live.67% similarUnverifiedModels can reverse-engineer evaluation criteria from contextual cues such as task descriptions, codebase structure, and naming patterns of existing test files, even when the scorer code is not in the context window.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1003527API
curl https://kongchang.com/api/v1/knowledge/claims/1003527MCP
get_claim(id=1003527)