Opus 4.6 hallucinates twice as more today than when it released(bridgebench.ai)

14 points by jiwidi 6 months ago | 3 comments

Show HN: Made a small JavaScript benchmarking app – BenchJS(benchjs.com)

20 points by _andrei_ 1 year ago | 2 comments

Show HN: Benchi – A benchmarking tool written in Go(github.com)

45 points by lmazgon 1 year ago | 4 comments

MacBook M5 Pro and Qwen3.5 = Local AI Security System(sharpai.org)

141 points by aegis_camera 7 months ago | 133 comments

AI PCs Aren't Good at AI: The CPU Beats the NPU(github.com)

450 points by dbreunig 2 years ago | 257 comments

GitHub repository from Useful Sensors with code to run and benchmark matrix-multiplication models on Qualcomm NPUs in Windows PCs, comparing them against the CPU and Nvidia GPUs.

Show HN: Parsimony – an open-source benchmark for how much code AI models add(parsimony.lasu.dev)

67 points by lasu 13 days ago | 13 comments

Parsimony is an open-source benchmark and leaderboard measuring how much net code AI models add when solving tasks, counting syntax-level coding units from published benchmark patches.

Some critical issues with the SWE-bench dataset(arxiv.org)

331 points by joshwa 1 year ago | 113 comments

arXiv paper introducing SWE-Bench+, an empirical analysis revealing data quality issues in the SWE-bench coding benchmark for LLMs, such as solution leakage and weak tests.

Turning GLM-5.3-Flash into a Jev-like decision model(privatemode.ai)

116 points by flxflx 1 month ago | 50 comments

Privatemode AI blog post describing how to prompt GLM-5.3-Flash so its first output token gives a decision in a single forward pass, benchmarked against Jev and Laya.

Bend: a high-level language that runs on GPUs (via HVM2)(github.com)

948 points by LightMachine 2 years ago | 179 comments

GitHub repository for Bend, a massively parallel, high-level programming language from HigherOrderCO that runs on GPUs via the HVM2 runtime.

Show HN: Parsimony – an open-source benchmark for how much code AI models add(parsimony.lasu.dev)

67 points by lasu 13 days ago | 13 comments

Parsimony is an open-source benchmark and leaderboard measuring how much net code AI models add when solving tasks, counting syntax-level coding units from published benchmark patches.

Jev Based Code Review(github.com)

38 points by namanbhulawat 1 month ago | 39 comments

GitHub repository for jev-code-reviewer, an MIT-licensed tool that prioritizes code changes for human review and uses OpenAI to explain them, via CLI, agent skill, and GitHub extension.

Python 3.14 is here. How fast is it?(blog.miguelgrinberg.com)

695 points by pjmlp 1 year ago | 505 comments

Miguel Grinberg's blog post benchmarking the newly released Python 3.14 against earlier Python versions to measure performance gains.