Evaluating AI's Scholarly Writing: Beyond Text Similarity
A new benchmark, RWGBench, redefines how we evaluate AI-generated related work by focusing on citation choices and scholarly positioning, rather than mere text similarity.
Read the full articleYou might also wanna read
PreScience: A Large-Scale Dataset and Benchmark for Forecasting Scientific Advances in AI Research
Can AI systems trained on the existing scientific record forecast the advances that will follow? We introduce PreScience, a dataset and benc
arxiv.org·6d ago
ParseBench — Document Parsing Benchmark for AI Agents
The first benchmark that evaluates what AI agents actually need from document parsing. 2,000 human-verified pages, 169K deterministic test r
parsebench.ai·15d ago
New benchmark reveals AI models often cite wrong sources even when answers are correct
Leading AI models like GPT and Gemini routinely cite text passages in document analyses that don't actually support their answers. Even when
the-decoder.com·1mo ago
SkillsBench: A Benchmark for Evaluating AI Agent Skills Across Diverse Tasks
Agent Skills are structured packages of procedural knowledge that augment LLM agents at inference time. Despite rapid adoption, there is no
arxiv.org·4mo ago
EdgeBench: A Benchmark for Measuring AI Environment Learning Through Extended Real-World Tasks
EdgeBench studies how agents learn from real-world environments across 134 day-long executable tasks.
edge-bench.org·7d ago
Benchmark Analysis: Comparing Document Parsing APIs for Enterprise AI Applications
Learn how Tensorlake built the most reliable document parsing API by measuring what actually matters: structural preservation, reading order
tensorlake.ai·8mo ago
Comments
Sign in to join the conversation.
No comments yet. Be the first.