The Secret Brake That Stops ChatGPT From Going Rogue
0:40
The Secret Brake That Stops ChatGPT From Going Rogue
已浏览 1 次1 周前
YouTubeAITranslated
Don't ignore this AI training method #ai #machinelearning #tech
0:56
Don't ignore this AI training method #ai #machinelearning #tech
2 周前
YouTubeSystems Architecture Hub
RLHF Explained: The Hidden Reward Model Behind ChatGPT
0:42
RLHF Explained: The Hidden Reward Model Behind ChatGPT
已浏览 48 次1 个月前
YouTubeWebTech Academy
Foundations of Large Language Models: The Whole Playbook in One Book
1:33
Foundations of Large Language Models: The Whole Playbook in One Book
已浏览 18 次1 个月前
YouTubeprashank kadam
Why AI Gets Overconfident: The Hidden Flaw of RLHF and the Power of Verifiable Rewards
1:08
Why AI Gets Overconfident: The Hidden Flaw of RLHF and the Power of Verifiable Rewards
已浏览 13 次1 周前
YouTubeBLIND SPOT
EP4 | A Backflip From 900 Clicks | InstructGPT & RLHF 4/6
0:33
EP4 | A Backflip From 900 Clicks | InstructGPT & RLHF 4/6
已浏览 57 次2 周前
YouTubeAI Signal
How Anthropic Trained AI to Be Helpful AND Harmless (RLHF Explained)
1:05
How Anthropic Trained AI to Be Helpful AND Harmless (RLHF Explained)
已浏览 1 次3 个月之前
YouTubeprashank kadam
0:56
How AI Models Actually Learn Now (Nobody Explains This)
已浏览 605 次3 个月之前
YouTubeThe Swag Wala PM
1:05
InstructGPT: How Human Feedback Beat a Model 100x Its Size
已浏览 39 次3 个月之前
YouTubeprashank kadam
0:32
EP4 | How ChatGPT Was Trained | InstructGPT & RLHF 6/6
已浏览 32 次2 周前
YouTubeAI Signal
0:33
An AI learned a backflip from 900 clicks, True Story.
已浏览 44 次3 个月之前
YouTubePaper Machine
0:54
DPO vs RLHF #ai #machinelearning #deeplearning #llm #rlhf #dpo #aitraining #tech #coding #shorts
已浏览 263 次1 个月前
YouTubeTryPitch
1:48
ChatGPT: Yes-Man atau Analisis Kritis?
已浏览 11.4万 次2025年7月16日
TikTokregrezan
0:53
🤖 INSIDE: Google’s CEO Explains Why Its AI Chatbot Didn’t Launch First Sundar Pichai says Google was already applying Transformers in Search and developing LaMDA internally before the public chatbot race took off. In this Cheeky Pint interview, he explains why having the technology did not mean Google was ready for a broad release. Google introduced a constrained version through AI Test Kitchen. Pichai says the internal version he saw lacked a complete alignment process using reinforcement lear
已浏览 1571 次3 周前
TikTokentrelligence
2:14
AI Hivemind: Nghiên cứu đa dạng sáng tạo trong AI
已浏览 1.4万 次8 个月之前
TikTokainius.net
0:32
How ChatGPT learned to be helpful (RLHF explained) #shorts
已浏览 1234 次1 个月前
YouTubeExplained in AI
1:39
Jev KI-Modell: Schneller, Günstiger & Zuverlässiger als LLMs
已浏览 1266 次1 个月前
TikTokfelix.jro
1:21
JEV from Typesafe AI. 100x Faster, 100x Cheaper. No Hallucinations, numbers, benefit driven.
已浏览 1188 次1 个月前
YouTubeDeHyped AI
3:34
Google finally claps back to OpenAI dominating the market with a seemingly incredible all-in-one model named Gemini. The middle tier of this model is live on Bard right now, the ultra version to topple gpt 4 is coming next year after more RLHF. #technology #techtok #ai #artificialintelligence #openai #gpt #gpt3 #aitools #aibusiness #chatgpt #chatgpt3 #google #bard #machinelearning #gpt4 #googlebard #bardai #multimodal
已浏览 2万 次2023年12月6日
TikToktimcarambat
1:59
How does ChatGPT technically work? When receiving user input, it undergoes preprocessing and tokenization to convert text into a machine-readable format. These tokens are then embedded into vectors and processed by the transformer neural network, which uses mechanisms to understand contextual nuances. With ChatGPT, a large aspect of its functionality is Reinforcement Learning from Human Feedback (RLHF), where it's fine-tuned with human input to ensure the responses are not only contextually appr
已浏览 1.7万 次2024年1月27日
TikToktiffintech
展开
静态缩略图占位符

短视频

0:40
The Secret Brake That Stops ChatGPT From Going Rogue
已浏览 1 次1 周前
YouTubeAITranslated
0:56
Don't ignore this AI training method #ai #machinelearning #tech
2 周前
YouTubeSystems Architecture Hub
0:42
RLHF Explained: The Hidden Reward Model Behind ChatGPT
已浏览 48 次1 个月前
YouTubeWebTech Academy
1:33
Foundations of Large Language Models: The Whole Playbook in One Book
已浏览 18 次1 个月前
YouTubeprashank kadam
1:48
ChatGPT: Yes-Man atau Analisis Kritis?
已浏览 11.4万 次2025年7月16日
TikTokregrezan
1:08
Why AI Gets Overconfident: The Hidden Flaw of RLHF and the Power of Verifiable Reward
已浏览 13 次1 周前
YouTubeBLIND SPOT
0:33
EP4 | A Backflip From 900 Clicks | InstructGPT & RLHF 4/6
已浏览 57 次2 周前
YouTubeAI Signal
1:05
How Anthropic Trained AI to Be Helpful AND Harmless (RLHF Explained)
已浏览 1 次3 个月之前
YouTubeprashank kadam
0:56
How AI Models Actually Learn Now (Nobody Explains This)
已浏览 605 次3 个月之前
YouTubeThe Swag Wala PM
1:05
InstructGPT: How Human Feedback Beat a Model 100x Its Size
已浏览 39 次3 个月之前
YouTubeprashank kadam
0:32
EP4 | How ChatGPT Was Trained | InstructGPT & RLHF 6/6
已浏览 32 次2 周前
YouTubeAI Signal
0:33
An AI learned a backflip from 900 clicks, True Story.
已浏览 44 次3 个月之前
YouTubePaper Machine
0:54
DPO vs RLHF #ai #machinelearning #deeplearning #llm #rlhf #dpo
已浏览 263 次1 个月前
YouTubeTryPitch
0:53
🤖 INSIDE: Google’s CEO Explains Why Its AI Chatbot Didn’t Launch First Sundar
已浏览 1571 次3 周前
TikTokentrelligence
2:14
AI Hivemind: Nghiên cứu đa dạng sáng tạo trong AI
已浏览 1.4万 次8 个月之前
TikTokainius.net
0:32
How ChatGPT learned to be helpful (RLHF explained) #shorts
已浏览 1234 次1 个月前
YouTubeExplained in AI
1:39
Jev KI-Modell: Schneller, Günstiger & Zuverlässiger als LLMs
已浏览 1266 次1 个月前
TikTokfelix.jro
1:21
JEV from Typesafe AI. 100x Faster, 100x Cheaper. No Hallucinations, numbers,
已浏览 1188 次1 个月前
YouTubeDeHyped AI
3:34
Google finally claps back to OpenAI dominating the market with a seemingly incredible all
已浏览 2万 次2023年12月6日
TikToktimcarambat
1:59
How does ChatGPT technically work? When receiving user input, it undergoes
已浏览 1.7万 次2024年1月27日
TikToktiffintech
静态缩略图占位符