Some of those that Codeforces…
December 26, 2024 at 9:38 PM
Some of those that Codeforces…
I Built an OCaml workspace for Codeforces. It has automated contest setup, problem scaffolding, and testing workflow for competitive programming. Perfect for beginners wanting structured development with FastIO optimization.
github.com/aguluman/codeforces
#OCaml #Codeforces #OpenSource
github.com/aguluman/codeforces
#OCaml #Codeforces #OpenSource
GitHub - aguluman/codeforces: OCaml Workspace for Codeforces Competitive Programming
OCaml Workspace for Codeforces Competitive Programming - aguluman/codeforces
github.com
September 13, 2025 at 4:09 AM
I Built an OCaml workspace for Codeforces. It has automated contest setup, problem scaffolding, and testing workflow for competitive programming. Perfect for beginners wanting structured development with FastIO optimization.
github.com/aguluman/codeforces
#OCaml #Codeforces #OpenSource
github.com/aguluman/codeforces
#OCaml #Codeforces #OpenSource
Gemini 3 deepthink has a 3455 rating on Codeforces... here are human ratings for comparison (only 7 humans above it)
#gamedev #gamedevelopment #indiedev #AI #gemini #codeforces #code #programming
#gamedev #gamedevelopment #indiedev #AI #gemini #codeforces #code #programming
February 13, 2026 at 12:03 PM
Gemini 3 deepthink has a 3455 rating on Codeforces... here are human ratings for comparison (only 7 humans above it)
#gamedev #gamedevelopment #indiedev #AI #gemini #codeforces #code #programming
#gamedev #gamedevelopment #indiedev #AI #gemini #codeforces #code #programming
DeepSeek releases DeepSeek-V4. 🐋
- DeepSeek-V4-Pro: 1.6T params
- DeepSeek-V4-Flash: 284B params
DeepSeek-V4-Pro rivals Claude-Opus-4.6-Max, GPT-5.4-xHigh and Gemini-3.1-Pro-High.
They support 1M context length, thinking and set new records for Codeforces.
- DeepSeek-V4-Pro: 1.6T params
- DeepSeek-V4-Flash: 284B params
DeepSeek-V4-Pro rivals Claude-Opus-4.6-Max, GPT-5.4-xHigh and Gemini-3.1-Pro-High.
They support 1M context length, thinking and set new records for Codeforces.
April 24, 2026 at 3:52 AM
DeepSeek releases DeepSeek-V4. 🐋
- DeepSeek-V4-Pro: 1.6T params
- DeepSeek-V4-Flash: 284B params
DeepSeek-V4-Pro rivals Claude-Opus-4.6-Max, GPT-5.4-xHigh and Gemini-3.1-Pro-High.
They support 1M context length, thinking and set new records for Codeforces.
- DeepSeek-V4-Pro: 1.6T params
- DeepSeek-V4-Flash: 284B params
DeepSeek-V4-Pro rivals Claude-Opus-4.6-Max, GPT-5.4-xHigh and Gemini-3.1-Pro-High.
They support 1M context length, thinking and set new records for Codeforces.
im sorry but too many huggingface models are shrimp brains overfit to codeforces problems
November 18, 2025 at 10:19 PM
im sorry but too many huggingface models are shrimp brains overfit to codeforces problems
yeah the worst part about working in the tech sector is that all the money is coming from Stanford philosophy grads who don't get that just because LLMs have high codeforces elos doesn't mean that self driving without LIDAR is possible suddenly
April 22, 2025 at 8:37 PM
yeah the worst part about working in the tech sector is that all the money is coming from Stanford philosophy grads who don't get that just because LLMs have high codeforces elos doesn't mean that self driving without LIDAR is possible suddenly
No known upper limit for RL
ProRL prolonged reinforcement learning (RL) for 3,000 steps. Far beyond what anyone else has done
and they didn't find the limits, the model kept getting better
also, a hot 1.5B open weights reasoning model in there
hijkzzz.notion.site/prorl-v2
ProRL prolonged reinforcement learning (RL) for 3,000 steps. Far beyond what anyone else has done
and they didn't find the limits, the model kept getting better
also, a hot 1.5B open weights reasoning model in there
hijkzzz.notion.site/prorl-v2
August 13, 2025 at 1:19 AM
No known upper limit for RL
ProRL prolonged reinforcement learning (RL) for 3,000 steps. Far beyond what anyone else has done
and they didn't find the limits, the model kept getting better
also, a hot 1.5B open weights reasoning model in there
hijkzzz.notion.site/prorl-v2
ProRL prolonged reinforcement learning (RL) for 3,000 steps. Far beyond what anyone else has done
and they didn't find the limits, the model kept getting better
also, a hot 1.5B open weights reasoning model in there
hijkzzz.notion.site/prorl-v2
Google's Deepthink has a 3455 elo rating on Codeforces lol
February 12, 2026 at 10:19 PM
Google's Deepthink has a 3455 elo rating on Codeforces lol
probably zero? until this moment I was not aware of codeforces 😆
but it seems like a fun idea!
roe.dev/ama
#ama
but it seems like a fun idea!
roe.dev/ama
#ama
May 8, 2026 at 8:26 AM
probably zero? until this moment I was not aware of codeforces 😆
but it seems like a fun idea!
roe.dev/ama
#ama
but it seems like a fun idea!
roe.dev/ama
#ama
Dijkstra is not optimal — proven by Codeforces legends - Codeforces
codeforces.com
June 3, 2025 at 9:52 AM
rawr 304
codeforces.com/blog/entry/1...
I am the poster child for CF rating deflation now due to AI apparently.
I am being violated by a data analysis (/j)
codeforces.com/blog/entry/1...
I am the poster child for CF rating deflation now due to AI apparently.
I am being violated by a data analysis (/j)
Codeforces rating deflation - Codeforces
codeforces.com
June 23, 2026 at 9:30 PM
rawr 304
codeforces.com/blog/entry/1...
I am the poster child for CF rating deflation now due to AI apparently.
I am being violated by a data analysis (/j)
codeforces.com/blog/entry/1...
I am the poster child for CF rating deflation now due to AI apparently.
I am being violated by a data analysis (/j)
I still prefer Claude Sonnet 3.5/3.6, but...
OpenAI's Competitive Programming with Large Reasoning Models
o3 achieves a gold medal at the 2024 IOI and obtains a Codeforces rating on par with elite human competitors.
arxiv.org/abs/2502.06807
OpenAI's Competitive Programming with Large Reasoning Models
o3 achieves a gold medal at the 2024 IOI and obtains a Codeforces rating on par with elite human competitors.
arxiv.org/abs/2502.06807
February 12, 2025 at 4:25 AM
I still prefer Claude Sonnet 3.5/3.6, but...
OpenAI's Competitive Programming with Large Reasoning Models
o3 achieves a gold medal at the 2024 IOI and obtains a Codeforces rating on par with elite human competitors.
arxiv.org/abs/2502.06807
OpenAI's Competitive Programming with Large Reasoning Models
o3 achieves a gold medal at the 2024 IOI and obtains a Codeforces rating on par with elite human competitors.
arxiv.org/abs/2502.06807
Secondly, every single MBPP test example and 78% of CodeForces have semantic duplicates (that is, some training data which are equivalent to items of the test set)
February 27, 2026 at 5:27 PM
Secondly, every single MBPP test example and 78% of CodeForces have semantic duplicates (that is, some training data which are equivalent to items of the test set)
Vayyy. GitHub, Cursor'a açtığı savaşta agresifleşiyor. Codeforces testinde haleflerinden %44 daha iyi olan o1, $10'dan başlayan (öğrencilere ücretsiz) paketine geliyor. Üstelik şu anda Cursor'ın $20'lık paketinde sadece "o1-mini" erişimi mevcut.
You can now use OpenAI’s new o1 model in GitHub Copilot Pro, Business, and Enterprise—and in GitHub Models. It’s all about that developer choice. 💻
github.blog/news-insight...
github.blog/news-insight...
OpenAI’s latest o1 model now available in GitHub Copilot and GitHub Models
OpenAI’s o1 model is now available in GitHub Copilot and GitHub Models, bringing advanced coding capabilities to your workflows.
github.blog
December 21, 2024 at 12:40 AM
Vayyy. GitHub, Cursor'a açtığı savaşta agresifleşiyor. Codeforces testinde haleflerinden %44 daha iyi olan o1, $10'dan başlayan (öğrencilere ücretsiz) paketine geliyor. Üstelik şu anda Cursor'ın $20'lık paketinde sadece "o1-mini" erişimi mevcut.
Kimi k1.5 --- an o1-level multi-modal model
- SOTA short-CoT performance, outperforming GPT-4o and Claude Sonnet 3.5 on 📐AIME, 📐MATH-500, 💻 LiveCodeBench by a large margin (up to +550%)
- Long-CoT performance matches o1 across multiple modalities (👀MathVista, 📐AIME, 💻Codeforces, etc)
Demo: kimi.ai
- SOTA short-CoT performance, outperforming GPT-4o and Claude Sonnet 3.5 on 📐AIME, 📐MATH-500, 💻 LiveCodeBench by a large margin (up to +550%)
- Long-CoT performance matches o1 across multiple modalities (👀MathVista, 📐AIME, 💻Codeforces, etc)
Demo: kimi.ai
January 20, 2025 at 4:49 PM
Kimi k1.5 --- an o1-level multi-modal model
- SOTA short-CoT performance, outperforming GPT-4o and Claude Sonnet 3.5 on 📐AIME, 📐MATH-500, 💻 LiveCodeBench by a large margin (up to +550%)
- Long-CoT performance matches o1 across multiple modalities (👀MathVista, 📐AIME, 💻Codeforces, etc)
Demo: kimi.ai
- SOTA short-CoT performance, outperforming GPT-4o and Claude Sonnet 3.5 on 📐AIME, 📐MATH-500, 💻 LiveCodeBench by a large margin (up to +550%)
- Long-CoT performance matches o1 across multiple modalities (👀MathVista, 📐AIME, 💻Codeforces, etc)
Demo: kimi.ai
Fiz meu primeiro contest no codeforces hoje, definitivamente uma experiência kkkkkkkk
February 9, 2025 at 6:41 PM
Fiz meu primeiro contest no codeforces hoje, definitivamente uma experiência kkkkkkkk
Всем привет! Завтра у меня муниципальный этап по информатике... Планы на сегодня: решить как можно больше муниципальных этапов на codeforces
December 14, 2024 at 9:11 AM
Всем привет! Завтра у меня муниципальный этап по информатике... Планы на сегодня: решить как можно больше муниципальных этапов на codeforces
codechef e codeforces são as que eu participo
July 18, 2026 at 3:35 AM
codechef e codeforces são as que eu participo
I recently made this plot for a talk I gave on AI progress and it helped me appreciate how quickly AI models are improving.
I know there's still a lot of benchmarks where progress is flat, but progress on Codeforces was quite flat for a long time too.
I know there's still a lot of benchmarks where progress is flat, but progress on Codeforces was quite flat for a long time too.
May 3, 2025 at 7:37 PM
I recently made this plot for a talk I gave on AI progress and it helped me appreciate how quickly AI models are improving.
I know there's still a lot of benchmarks where progress is flat, but progress on Codeforces was quite flat for a long time too.
I know there's still a lot of benchmarks where progress is flat, but progress on Codeforces was quite flat for a long time too.
🚀 CF Ladder – Smarter Way to Practice Codeforces Problems Hey everyone 👋 I’m super excited to share CF Ladder, a web app I built to make practicing Codeforces problems more organized, eff...
#webdev #react #nextjs #frontend
Origin | Interest | Match
#webdev #react #nextjs #frontend
Origin | Interest | Match
🚀 CF Ladder – Smarter Way to Practice Codeforces Problems
Hey everyone 👋 I’m super excited to share CF Ladder, a web app I built to make practicing Codeforces...
dev.to
October 31, 2025 at 7:52 AM