arXiv AI By Ziyue Wang (State Key Laboratory of Multimedia Information Processing, School of Computer Science, Peking University), Aomufei Yuan (Peking University), Yiran Yao (Tianjin University), Linli Yao (State Key Laboratory of Multimedia Information Processing, School of Computer Science, Peking University), Hongyao Zuo (Tianjin University), Ziwen Gong (Hainan University), Yuanxin Liu (State Key Laboratory of Multimedia Information Processing, School of Computer Science, Peking University), Shicheng Li (State Key Laboratory of Multimedia Information Processing, School of Computer Science, Peking University), Yishuo Cai (State Key Laboratory of Multimedia Information Processing, School of Computer Science, Peking University), Tong Yang (Peking University), Xu Sun (State Key Laboratory of Multimedia Information Processing, School of Computer Science, Peking University), Xiaohui Li (Huawei Technologies), Haoli Bai (Huawei Technologies)

What Proves You Wrong: Benchmarking Language Models on Falsifiable Research Ideation

Read the original on arXiv AI →

The paper introduces Lit2Test, a benchmark that evaluates language models’ research idea proposals by requiring each idea to include a falsifiable outcome, thereby making quality decidable. Built from 200 real-paper neighborhoods, the benchmark gathers proposals from four frontier models and compares them via 1,200 blind pairwise judgments, with reliability checks and human calibration. The results show a consistent ranking of the models, driven by test and metric quality rather than fluency, and the authors release the benchmark and related artifacts for public use.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

arXiv AI
Jul 7

Code Benchmarks Should Prioritize Rigor, Reliability, and Reproducibility

arXiv:2501. 10711v5 Announce Type: replace-cross Abstract: Code-related benchmarks play a critical role in evaluating large language models (LLMs), yet their quality fundamentally shapes how the community interprets model capabilities.

By Jialun Cao, Yuk-Kit Chan, Zixuan Ling, Wenxuan Wang, Shuqing Li, Mingwei Liu, Ruixi Qiao, Yuting Han, Chaozheng Wang, Boxi Yu, Pinjia He, Shuai Wang, Zibin Zheng, Michael R. Lyu, Shing-Chi Cheung
arXiv Computation and Language
Sep 1

When Calibration Rankings Reverse: Accuracy-Controlled Evaluation for Fair Comparison of LLMs

The paper argues that traditional global calibration metrics, such as Expected Calibration Error and Brier Score, are confounded by differences in model accuracy when comparing large language models. It introduces ACE, an accuracy‑controlled evaluation framework that offers Instance‑Aligned, Distribution‑Aligned, and Candidate‑Aligned views to provide fairer cross‑model comparisons. Experiments across various benchmarks reveal that many reported calibration advantages disappear after accuracy control and that model rankings often reverse, indicating that raw global metrics are unreliable for cross‑model calibration assessment.

By Zhichao Yang, Caiqi Zhang, Ruihan Yang, Chengzu Li, Nigel Collier, Deqing Yang
arXiv AI
Sep 3

EvalDetectBench: A Benchmark for Measuring Evaluation Awareness in Frontier Language Models

EvalDetectBench is an open pipeline and benchmark designed to measure evaluation awareness in frontier large language models, enabling practitioners to test models against any Inspect-compatible evaluation. It includes a curated transcript suite from current frontier system-card evaluations and diverse deployment sources, and it assesses both how reliably models recognize they are being evaluated and how detectable individual benchmarks are. The benchmark addresses systematic bias by calibrating probes per model and harmonizing generator selection to correct for variance caused by model identity and prompt choice.

By Xinning Li, Kemunto Ochwang'i, Aryasomayajula Ram Bharadwaj, Alexandra Souly, Robert Kirk