User contributions for 117.195.140.85

18 April 2025

• 07:0307:03, 18 April 2025 diff hist +1 Humanity's Last Exam →Results: uh oh, I see why it's difficult to update the date Tags: Mobile edit Mobile web edit
• 01:4801:48, 18 April 2025 diff hist +42 Humanity's Last Exam →Creation: minor fragment add Tags: Mobile edit Mobile web edit

17 April 2025

• 23:0223:02, 17 April 2025 diff hist +13 Humanity's Last Exam →Results: Experimental. apparently important, "Preview" behaves differently, and the "actual release" may yet different. I have not removed o3-mini here, to give company and a comparison point to the also text-only R1, the best performing open-weight model Tags: Mobile edit Mobile web edit
• 21:4521:45, 17 April 2025 diff hist +71 Humanity's Last Exam lead Tags: Mobile edit Mobile web edit
• 21:3221:32, 17 April 2025 diff hist +31 Humanity's Last Exam →Creation: bug bounty Tags: Mobile edit Mobile web edit
• 17:3417:34, 17 April 2025 diff hist −2 Humanity's Last Exam →Results: updated the text only numbers. the variability in the "calibration error" numbers is confusing. is the "calib" in SEAL Leaderboards even the same thing as the calibration error on the HLE main page? Tags: Mobile edit Mobile web edit
• 17:3017:30, 17 April 2025 diff hist +8 Humanity's Last Exam →Results: replace o1 results with o3 (high), use SEAL leaderboards instead of HLE homepage (they didn't even update the "last updated" date after adding the new models...). will update the text only numbers in the next edit Tags: Mobile edit Mobile web edit

3 March 2011

• 12:4712:47, 3 March 2011 diff hist +19 TI-84 Plus series No edit summary
• 12:4412:44, 3 March 2011 diff hist +19 Wikipedia:Sandbox No edit summary

12 March 2010

• 10:0510:05, 12 March 2010 diff hist +6 Tirumalagiri →Restaurants