HN user

francesco

4 karma
Posts1
Comments6
View on HN

It looks to me like the performance reported for rStar-Math (both, in the table and in the graph) is incorrect. With a single rollout (at test time) rStar-Math achieves 50 in AIME 2024 not 26.7 as you reported. In Olympiad Bench it achieves 65.3, not 47.1 as you reported. In AMC 2023 it achieves 87.5, not 47.5 as your reported. It outperforms your model across the board. Am I reading something incorrectly?

Unlike Gillette, safety razors do not use proprietary systems for their blades. The consequence is you can use any double edge blade and you can have fun finding the double edge blades that best matches your razor and beard. Shameless plug: years ago with my brother I made this small site (https://www.razorbladesclub.com) that I presented on "Show HN" exactly to allow people to buy samplers of blades and try them out. It didn't get much attention then but years later the site is still alive and hopefully it still serves people well :)

Author here. I and my brother have been working on this for the past couple of months and would love to get some feedback on it. It's a small service for the wet-shaving community aiming at offering a variety of double edge blades (even in quantities as low as one blade) as a one-off purchase or as a subscription (which took a long time to implement:)). I don't know whether the message is clear, whether the idea could work and if there are things that we could improve (of course there are :)) Any suggestions or feedback will be greatly appreciated. Thank you very much.

An HTTP post & response could be an option. It would just open the possibility for other services to integrate with yours, there making it more flexible and open to a larger audience. A freemium approach could then be a possibility (where one starts paying after a certain amount of daily API usage)