pull down to refresh

FrontierMath: A Benchmark for Evaluating Advanced Mathematical Reasoning in AIepochai.org/frontiermath/the-benchmark
203 sats \ 0 comments \ @Rsync25 10 Nov 2024 tech
related
Testing AI systems on hard math problems shows they still perform very poorlyphys.org/news/2024-11-ai-hard-math-problems-poorly.html
153 sats \ 4 comments \ @south_korea_ln 13 Nov 2024 science
AI is actually bad at math, ORCA showswww.theregister.com/2025/11/17/ai_bad_math_orca/
167 sats \ 4 comments \ @0xbitcoiner 18 Nov AI
DeepPHY: Benchmarking Agentic VLMs on Physical Reasoningarxiv.org/abs/2508.05405
182 sats \ 0 comments \ @optimism 10 Aug AI
AI benchmarks hampered by bad sciencewww.theregister.com/2025/11/07/measuring_ai_models_hampered_by/
178 sats \ 0 comments \ @0xbitcoiner 10 Nov AI
AI mimics neocortex computations with 'winner-take-all' approachtechxplore.com/news/2024-10-ai-mimics-neocortex-winner-approach.html
49 sats \ 0 comments \ @ch0k1 26 Oct 2024 tech
Scaling up: how increasing inputs has made artificial intelligence more capableourworldindata.org/scaling-up-ai
279 sats \ 0 comments \ @0xbitcoiner 20 Jan charts_and_numbers
The AI Industry's Scaling Obsession Is Headed for a Cliffarxiv.org/abs/2507.07931
321 sats \ 9 comments \ @0xbitcoiner 15 Oct AI
OpenAI o3 beats FrontierMath - because OpenAI funded the test and cheatedpivot-to-ai.com/2025/01/20/openai-o3-beats-frontiermath-because-openai-funded-the-test-and-had-access-to-questions/
366 sats \ 1 comment \ @StillStackinAfterAllTheseYears 21 Jan tech
Debate May Help AI Models Converge on Truthwww.quantamagazine.org/debate-may-help-ai-models-converge-on-truth-20241108/
237 sats \ 0 comments \ @0xbitcoiner 8 Nov 2024 science
Scaling up: how increasing inputs has made artificial intelligence more capable
215 sats \ 0 comments \ @zuspotirko 29 Jan tech
Olympiad-level formal mathematical reasoning with reinforcement learningwww.nature.com/articles/s41586-025-09833-y
175 sats \ 2 comments \ @0xbitcoiner 20 Nov AI
AI achieves silver-medal standard solving International Mathematical Olympiaddeepmind.google/discover/blog/ai-solves-imo-problems-at-silver-medal-level/
284 sats \ 3 comments \ @sancristrader 25 Jul 2024 tech
DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open LLMsarxiv.org/abs/2402.03300
662 sats \ 0 comments \ @zuspotirko 6 Feb 2024 science
Inside the Secret Meeting Where Mathematicians Struggled to Outsmart AIwww.scientificamerican.com/article/inside-the-secret-meeting-where-mathematicians-struggled-to-outsmart-ai/
421 sats \ 5 comments \ @k00b 9 Jun science
New MIT Research Proves AGI Was Achievedwww.geeky-gadgets.com/artificial-general-intelligence-advancements/
65 sats \ 1 comment \ @ch0k1 16 Nov 2024 news
Hermes 4 outperforms OpenAI models with minimal content restrictions
121 sats \ 0 comments \ @lunin 1 Sep AI
BrokenMath: A Benchmark for Sycophancy in Theorem Proving with LLMsarxiv.org/abs/2510.04721
180 sats \ 1 comment \ @jakoyoh629 25 Oct AI
Quantitative AI progress needs accurate and transparent evaluationmathstodon.xyz/@tao/114910028356641733
100 sats \ 0 comments \ @hn 25 Jul tech
Google's DeepMind AI decodes age-old math equation, stumping humansinterestingengineering.com/innovation/googles-deepmind-decodes-old-math-equation
100 sats \ 0 comments \ @zuspotirko 16 Dec 2023 AI
Chinese researchers just built an open-source rival to ChatGPT in 2 monthswww.livescience.com/technology/artificial-intelligence/china-releases-a-cheap-open-rival-to-chatgpt-thrilling-some-scientists-and-panicking-silicon-valley
21 sats \ 0 comments \ @ch0k1 25 Jan news
Researchers isolate memorization from problem-solving in AI neural networksarstechnica.com/ai/2025/11/study-finds-ai-models-store-memories-and-logic-in-different-neural-regions/
400 sats \ 1 comment \ @0xbitcoiner 11 Nov AI