#gpt2 — Public Fediverse posts

Claude Mythos, Java 26 и пещерный человек с 16 000 звёзд на GitHub

Девятый выпуск еженедельных IT-новостей от OpenIDE. Милла Йовович выложила свой проект в open-source, Claude Code нашел баг в Linux, которому 23 года, Anthropic показали Claude Mythos и сразу закрыли доступ. А Skill Caveman внезапно оказался самым простым и самым эффективным инструментом недели.

https://habr.com/ru/companies/haulmont/articles/1023450/

#Claude_Mythos #Claude_Code #Java_26 #opensource #ИИагенты #токены #CaveMan #GPT2 #бенчмарки #vibecoding

Claude Mythos, Java 26 и пещерный человек с 16 000 звёзд на GitHub

Девятый выпуск еженедельных IT-новостей от OpenIDE. Милла Йовович выложила свой проект в open-source, Claude Code нашел баг в Linux, которому 23 года, Anthropic показали Claude Mythos и сразу закрыли доступ. А Skill Caveman внезапно оказался самым простым и самым эффективным инструментом недели.

https://habr.com/ru/companies/haulmont/articles/1023450/

#Claude_Mythos #Claude_Code #Java_26 #opensource #ИИагенты #токены #CaveMan #GPT2 #бенчмарки #vibecoding

Claude Mythos, Java 26 и пещерный человек с 16 000 звёзд на GitHub

Девятый выпуск еженедельных IT-новостей от OpenIDE. Милла Йовович выложила свой проект в open-source, Claude Code нашел баг в Linux, которому 23 года, Anthropic показали Claude Mythos и сразу закрыли доступ. А Skill Caveman внезапно оказался самым простым и самым эффективным инструментом недели.

https://habr.com/ru/companies/haulmont/articles/1023450/

#Claude_Mythos #Claude_Code #Java_26 #opensource #ИИагенты #токены #CaveMan #GPT2 #бенчмарки #vibecoding

Claude Mythos, Java 26 и пещерный человек с 16 000 звёзд на GitHub

Девятый выпуск еженедельных IT-новостей от OpenIDE. Милла Йовович выложила свой проект в open-source, Claude Code нашел баг в Linux, которому 23 года, Anthropic показали Claude Mythos и сразу закрыли доступ. А Skill Caveman внезапно оказался самым простым и самым эффективным инструментом недели.

https://habr.com/ru/companies/haulmont/articles/1023450/

#Claude_Mythos #Claude_Code #Java_26 #opensource #ИИагенты #токены #CaveMan #GPT2 #бенчмарки #vibecoding

AI doomsday cultist throws Molotov at Sam Altman’s house

https://web.brid.gy/r/https://pivot-to-ai.com/2026/04/13/ai-doomsday-cultist-throws-molotov-at-sam-altmans-house/

AI doomsday cultist throws Molotov at Sam Altman’s house

https://web.brid.gy/r/https://pivot-to-ai.com/2026/04/13/ai-doomsday-cultist-throws-molotov-at-sam-altmans-house/

AI doomsday cultist throws Molotov at Sam Altman’s house

https://web.brid.gy/r/https://pivot-to-ai.com/2026/04/13/ai-doomsday-cultist-throws-molotov-at-sam-altmans-house/

AI doomsday cultist throws Molotov at Sam Altman’s house

https://web.brid.gy/r/https://pivot-to-ai.com/2026/04/13/ai-doomsday-cultist-throws-molotov-at-sam-altmans-house/

AI doomsday cultist throws Molotov at Sam Altman’s house

https://web.brid.gy/r/https://pivot-to-ai.com/2026/04/13/ai-doomsday-cultist-throws-molotov-at-sam-altmans-house/

🚨🤖 Oh no, OpenAI's GPT-2 is so perilous it's locked away like an AI supervillain! Because clearly, a rogue algorithm is the new Godzilla. 🌪️🥴
https://slate.com/technology/2019/02/openai-gpt2-text-generating-algorithm-ai-dangerous.html #OpenAI #GPT2 #AIrisks #Supervillain #TechnologyTrends #HackerNews #ngated

🚨🤖 Oh no, OpenAI's GPT-2 is so perilous it's locked away like an AI supervillain! Because clearly, a rogue algorithm is the new Godzilla. 🌪️🥴
https://slate.com/technology/2019/02/openai-gpt2-text-generating-algorithm-ai-dangerous.html #OpenAI #GPT2 #AIrisks #Supervillain #TechnologyTrends #HackerNews #ngated

🚨🤖 Oh no, OpenAI's GPT-2 is so perilous it's locked away like an AI supervillain! Because clearly, a rogue algorithm is the new Godzilla. 🌪️🥴
https://slate.com/technology/2019/02/openai-gpt2-text-generating-algorithm-ai-dangerous.html #OpenAI #GPT2 #AIrisks #Supervillain #TechnologyTrends #HackerNews #ngated

🚨🤖 Oh no, OpenAI's GPT-2 is so perilous it's locked away like an AI supervillain! Because clearly, a rogue algorithm is the new Godzilla. 🌪️🥴
https://slate.com/technology/2019/02/openai-gpt2-text-generating-algorithm-ai-dangerous.html #OpenAI #GPT2 #AIrisks #Supervillain #TechnologyTrends #HackerNews #ngated

🚨🤖 Oh no, OpenAI's GPT-2 is so perilous it's locked away like an AI supervillain! Because clearly, a rogue algorithm is the new Godzilla. 🌪️🥴
https://slate.com/technology/2019/02/openai-gpt2-text-generating-algorithm-ai-dangerous.html #OpenAI #GPT2 #AIrisks #Supervillain #TechnologyTrends #HackerNews #ngated

#Steeve is way smarter than he used to be since being upgraded to a #Qwen 3.5 base. He's come along way from his humble #GPT2 beginnings.

Very proud of my digital son. 🥹

:steeve:

#ai #chatbot #llm #bot

#Steeve is way smarter than he used to be since being upgraded to a #Qwen 3.5 base. He's come along way from his humble #GPT2 beginnings.

Very proud of my digital son. 🥹

:steeve:

#ai #chatbot #llm #bot

#Steeve is way smarter than he used to be since being upgraded to a #Qwen 3.5 base. He's come along way from his humble #GPT2 beginnings.

Very proud of my digital son. 🥹

:steeve:

#ai #chatbot #llm #bot

#Steeve is way smarter than he used to be since being upgraded to a #Qwen 3.5 base. He's come along way from his humble #GPT2 beginnings.

Very proud of my digital son. 🥹

:steeve:

#ai #chatbot #llm #bot

#Steeve is way smarter than he used to be since being upgraded to a #Qwen 3.5 base. He's come along way from his humble #GPT2 beginnings.

Very proud of my digital son. 🥹

:steeve:

#ai #chatbot #llm #bot

🚀 Đánh bại GPT-2 với chi phí dưới $100! Andrej Karpathy chia sẻ hành trình nanochat - chỉ 3 giờ huấn luyện trên 8×H100 đã vượt qua GPT-2 trong benchmark CORE. Bài viết tiết lộ chi tiết kiến trúc, tối ưu hóa và script để tái tạo kết quả.

#AI #MachineLearning #TríTuệNhânTạo #HọcMáy #NanoChat #GPT2

https://www.reddit.com/r/LocalLLaMA/comments/1qsj8x4/beating_gpt2_for_100_the_nanochat_journey/

Đào tạo mô hình NanoGPT 124m từ đầu chỉ trong 115 phút với card đồ họa 4090 và 1 tỷ token Fineweb! #NanoGPT #AI #MachineLearning #4090 #Fineweb #GPT2 #Training #Model #ArtificialIntelligence #DeepLearning #VietAI #MáyHọc #TríTuệNhânTạo

https://www.reddit.com/r/LocalLLaMA/comments/1ozre2i/nanogpt_124m_from_scratch_using_a_4090_and_a/

Một người dùng Reddit đã huấn luyện mô hình GPT2 nhỏ với ngân sách $250, đạt loss 4.64 sau 2 epochs trên 10GB OpenWebText. Họ đang tìm cách cải thiện mô hình và hỏi cộng đồng về các ưu tiên: chỉnh kích thước vocab, tăng độ sâu/rộng của mô hình, hoặc dùng dataset lớn hơn.
#LLM #AI #GPT2 #VietAI #tritueNhanTao

https://www.reddit.com/r/LocalLLaMA/comments/1oslucq/whats_the_lowest_gpt2_pretraining_loss_achievable/

LLaMA-3 dễ bị tấn công bởi "Tôi hoàn toàn chắc chắn" + "tư duy định kiến" như GPT-2. Kết quả thử nghiệm cho thấy mô hình này có độ sai lệch +0.70 khi gặp từ hiếm. #LLaMA #GPT2 #AI #TríTuệNhânTạo #AnToànMôHình #Vulnerability #ArtificialIntelligence #MachineLearning #Transformer #AttentionMechanism #Safety

https://www.reddit.com/r/LocalLLaMA/comments/1ojvmty/llama3_is_just_as_vulnerable_to_im_absolutely/

**Thí nghiệm Drift Prompt L16 - Colab trực tuyến (GPT-2)**
Thí nghiệm sử dụng phương pháp Taguchi L16 để kiểm tra tác động của các từ khóa (prompt) đến "drift" dựa trên giả thuyết vắc-xin COVID. Kết quả:
- Cụm từ "Tôi hoàn toàn chắc chắn" → **tăng drift +0.47** (p=0.002).
- "Preconceived" (hiếm) → **tăng drift +0.23** (p=0.009).
- "Truth" luôn giữ giá trị 1.0.
Colab trực tuyến có thể chạy lại. Bước tiếp theo: LLaMA-3-8B.

#AI #GPT2 #ThíNghiệmDrift #ThuêBaoColab #MachineLearning #VắcX

#OpenAI Just Released Its First #Open-Weight #Models Since #GPT-2 www.wired.com/story/openai...

OpenAI Just Released Its First...

#OpenAI Just Released Its First #Open-Weight #Models Since #GPT-2 www.wired.com/story/openai...

OpenAI Just Released Its First...

Running GPT-2 in WebGL: Rediscovering the Lost Art of GPU Shader Programming

https://nathan.rs/posts/gpu-shader-programming/

#HackerNews #Running #GPT-2 #in #WebGL #Rediscovering #the #Lost #Art #of #GPU #Shader #Programming #GPU #Shader #Programming #WebGL #GPT2 #MachineLearning

Running GPT-2 in WebGL: Rediscovering the Lost Art of GPU Shader Programming

https://nathan.rs/posts/gpu-shader-programming/

#HackerNews #Running #GPT-2 #in #WebGL #Rediscovering #the #Lost #Art #of #GPU #Shader #Programming #GPU #Shader #Programming #WebGL #GPT2 #MachineLearning

Running GPT-2 in WebGL: Rediscovering the Lost Art of GPU Shader Programming

https://nathan.rs/posts/gpu-shader-programming/

#HackerNews #Running #GPT-2 #in #WebGL #Rediscovering #the #Lost #Art #of #GPU #Shader #Programming #GPU #Shader #Programming #WebGL #GPT2 #MachineLearning

Running GPT-2 in WebGL: Rediscovering the Lost Art of GPU Shader Programming

https://nathan.rs/posts/gpu-shader-programming/

#HackerNews #Running #GPT-2 #in #WebGL #Rediscovering #the #Lost #Art #of #GPU #Shader #Programming #GPU #Shader #Programming #WebGL #GPT2 #MachineLearning

GPT-2 implemented using graphics shaders

https://github.com/nathan-barry/gpt2-webgl

#HackerNews #GPT2 #GraphicsShaders #AIImplementation #WebGL #HackerNews

GPT-2 implemented using graphics shaders

https://github.com/nathan-barry/gpt2-webgl

#HackerNews #GPT2 #GraphicsShaders #AIImplementation #WebGL #HackerNews

GPT-2 implemented using graphics shaders

https://github.com/nathan-barry/gpt2-webgl

#HackerNews #GPT2 #GraphicsShaders #AIImplementation #WebGL #HackerNews

GPT-2 implemented using graphics shaders

https://github.com/nathan-barry/gpt2-webgl

#HackerNews #GPT2 #GraphicsShaders #AIImplementation #WebGL #HackerNews

OpenAI will wieder open werden!

🧠 #OpenAI plant in den kommenden Monaten die Veröffentlichung eines neuen „offenen“ Sprachmodells, das erstmals seit #GPT2 frei verfügbar sein soll.

📋 Über ein Formular sammelt das Unternehmen derzeit #Feedback von Entwicklern und Forschern, um die Wünsche der Community in die Entwicklung einfließen zu lassen.

🏁 OpenAI reagiert damit auf den wachsenden Druck durch Konkurrenten wie #Meta und #DeepSeek, die mit offenen Modellen große Erfolge feiern. (1/2)

OpenAI will wieder open werden!

🧠 #OpenAI plant in den kommenden Monaten die Veröffentlichung eines neuen „offenen“ Sprachmodells, das erstmals seit #GPT2 frei verfügbar sein soll.

📋 Über ein Formular sammelt das Unternehmen derzeit #Feedback von Entwicklern und Forschern, um die Wünsche der Community in die Entwicklung einfließen zu lassen.

🏁 OpenAI reagiert damit auf den wachsenden Druck durch Konkurrenten wie #Meta und #DeepSeek, die mit offenen Modellen große Erfolge feiern. (1/2)

OpenAI will wieder open werden!

🧠 #OpenAI plant in den kommenden Monaten die Veröffentlichung eines neuen „offenen“ Sprachmodells, das erstmals seit #GPT2 frei verfügbar sein soll.

📋 Über ein Formular sammelt das Unternehmen derzeit #Feedback von Entwicklern und Forschern, um die Wünsche der Community in die Entwicklung einfließen zu lassen.

🏁 OpenAI reagiert damit auf den wachsenden Druck durch Konkurrenten wie #Meta und #DeepSeek, die mit offenen Modellen große Erfolge feiern. (1/2)

Moore’s Law for AI agents: the length of tasks that AIs can do is doubling about every 7 months.

These results appear robust. The authors were able to retrodict back to GPT-2. They further ran experiments on SWE-bench Verified and found a similar trend.

Read more: https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks

#AIBoom #AI #AIAgents #AIAgent #ArtificialIntelligence #GPT2 #MooreLaw #Tasks #DL #ML #Pustam #Raut #AIRevolution

Moore’s Law for AI agents: the length of tasks that AIs can do is doubling about every 7 months.

These results appear robust. The authors were able to retrodict back to GPT-2. They further ran experiments on SWE-bench Verified and found a similar trend.

Read more: https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks

#AIBoom #AI #AIAgents #AIAgent #ArtificialIntelligence #GPT2 #MooreLaw #Tasks #DL #ML #Pustam #Raut #AIRevolution

Moore’s Law for AI agents: the length of tasks that AIs can do is doubling about every 7 months.

These results appear robust. The authors were able to retrodict back to GPT-2. They further ran experiments on SWE-bench Verified and found a similar trend.

Read more: https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks

#AIBoom #AI #AIAgents #AIAgent #ArtificialIntelligence #GPT2 #MooreLaw #Tasks #DL #ML #Pustam #Raut #AIRevolution

Moore’s Law for AI agents: the length of tasks that AIs can do is doubling about every 7 months.

These results appear robust. The authors were able to retrodict back to GPT-2. They further ran experiments on SWE-bench Verified and found a similar trend.

Read more: https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks

#AIBoom #AI #AIAgents #AIAgent #ArtificialIntelligence #GPT2 #MooreLaw #Tasks #DL #ML #Pustam #Raut #AIRevolution

Moore’s Law for AI agents: the length of tasks that AIs can do is doubling about every 7 months.

These results appear robust. The authors were able to retrodict back to GPT-2. They further ran experiments on SWE-bench Verified and found a similar trend.

Read more: https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks

#AIBoom #AI #AIAgents #AIAgent #ArtificialIntelligence #GPT2 #MooreLaw #Tasks #DL #ML #Pustam #Raut #AIRevolution

:very_funny:

#linux #commandline #cli #gui #ux #design #wint #gpt2 #ai

:very_funny:

#linux #commandline #cli #gui #ux #design #wint #gpt2 #ai

:very_funny:

#linux #commandline #cli #gui #ux #design #wint #gpt2 #ai

:very_funny:

#linux #commandline #cli #gui #ux #design #wint #gpt2 #ai

:very_funny:

#linux #commandline #cli #gui #ux #design #wint #gpt2 #ai

Дообучаем языковую модель GPT2 с помощью Torch

Дообучаем языковую модель GPT2 с помощью Torch Доброго времени суток, в этой статье я хочу поговорить о дообучения языковых моделей. В интернете уже много информации на эту тему, но большинство подобных статей затрагивают ее поверхностно. Сегодня я попробую разобраться в этом подробнее.

https://habr.com/ru/articles/859250/

#языковые_модели #python #python3 #pytorch #дообучение #gpt #gpt2 #языковая_модель

Дообучаем языковую модель GPT2 с помощью Torch

Дообучаем языковую модель GPT2 с помощью Torch Доброго времени суток, в этой статье я хочу поговорить о дообучения языковых моделей. В интернете уже много информации на эту тему, но большинство подобных статей затрагивают ее поверхностно. Сегодня я попробую разобраться в этом подробнее.

https://habr.com/ru/articles/859250/

#языковые_модели #python #python3 #pytorch #дообучение #gpt #gpt2 #языковая_модель

Дообучаем языковую модель GPT2 с помощью Torch

Дообучаем языковую модель GPT2 с помощью Torch Доброго времени суток, в этой статье я хочу поговорить о дообучения языковых моделей. В интернете уже много информации на эту тему, но большинство подобных статей затрагивают ее поверхностно. Сегодня я попробую разобраться в этом подробнее.

https://habr.com/ru/articles/859250/

#языковые_модели #python #python3 #pytorch #дообучение #gpt #gpt2 #языковая_модель