Even GPT-5.2 Can't Count to Five: Zero-Error Horizons in Trustworthy LLMs
Article URL: [https://arxiv.org/abs/2601.15714][1]
Comments URL: [https://news.ycombinator.com/item?id=47615876][2]
Points: 7
# Comments: 2
[1]: https://arxiv.org/abs/2601.15714
[2]: https://news.ycombinator.com/item?id=47615876
https://arxiv.org/abs/2601.15714