LargeLanguageModels

Chris Manning: Language Is the Real Unlock for Intelligence

Chris Manning is a legend in NLP. He's a professor of linguistics and computer science at Stanford and ran the Stanford AI Lab.

55:51

Chris Manning: Language Is the Real Unlock for Intelligence

8 views

44 min ago

Mehdi Hosseini Moghadam

L-39: In-context learning emerges – Few-Shot vs Zero-Shot Accuracy #LLM #Agents

L-39: In-context learning emerges (Medium) This video explores the concept of in-context learning, a phenomenon where large ...

4:27

L-39: In-context learning emerges – Few-Shot vs Zero-Shot Accuracy #LLM #Agents

0 views

19 min ago

Machinematics

The Architecture Shift That Built Modern LLMs

Why did decoder-only architectures (such as GPT, Llama, and modern frontier models) completely dominate the LLM landscape, ...

10:51

The Architecture Shift That Built Modern LLMs

2 views

15 min ago

Mehdi Hosseini Moghadam

L-32: Where training data comes from – LLM Data Curation #LLM #DataScience

L-32: Where training data comes from (Medium) This video addresses the critical challenge of data curation for large language ...

4:23

L-32: Where training data comes from – LLM Data Curation #LLM #DataScience

0 views

26 min ago

DevDrop

What an AI Model Actually Is (Explained Simply, No Maths) | AI Basics 1

Lesson 1 of DevDrop Basics: AI from zero to advanced, one short lesson at a time. If you've used ChatGPT or any AI assistant and ...

7:51

What an AI Model Actually Is (Explained Simply, No Maths) | AI Basics 1

0 views

19 min ago

Mehdi Hosseini Moghadam

L-43: Constrained decoding and structured output – Constrained Decoding #LLM #Agents #JSON

L-43: Constrained decoding and structured output (Medium) This video addresses the common issue of large language models ...

3:40

L-43: Constrained decoding and structured output – Constrained Decoding #LLM #Agents #JSON

0 views

16 min ago

Mehdi Hosseini Moghadam

L-35: Mixed precision: fp32, fp16, bf16 – Mixed Precision Training #llm #deeplearning

L-35: Mixed precision: fp32, fp16, bf16 (Medium) This video addresses the challenge of training large language models efficiently ...

3:21

L-35: Mixed precision: fp32, fp16, bf16 – Mixed Precision Training #llm #deeplearning

0 views

23 min ago

Skill MD

This AI "Can't Hallucinate." Its Makers Say That Number Isn't Empirical.

TypeSafe AI's launch post says its model Jev "can't hallucinate." A few paragraphs later: "Our number is not empirical. Schema ...

5:29

This AI "Can't Hallucinate." Its Makers Say That Number Isn't Empirical.

0 views

51 min ago

The Employee Factory

Stop waiting on slow LLMs! Discover how AnyJev-193X delivers lightning-fast AI performance that outpaces traditional Large ...

8:01

AnyJev-193X: Faster Than LLMs?

0 views

51 min ago

Mehdi Hosseini Moghadam

L-41: Beam search and its failures – Beam Search Implementation and Failures #llm #beamsearch #ai

L-41: Beam search and its failures (Medium) This video explores beam search, a decoding strategy used in large language ...

3:19

L-41: Beam search and its failures – Beam Search Implementation and Failures #llm #beamsearch #ai

0 views

17 min ago

Mehdi Hosseini Moghadam

L-28: Assemble a transformer block – Attention, FFN & Residuals #transformer #llm

L-28: Assemble a transformer block (Medium) This video addresses a core component of modern large language models: the ...

3:32

L-28: Assemble a transformer block – Attention, FFN & Residuals #transformer #llm

0 views

30 min ago

Mehdi Hosseini Moghadam

L-37: Parallelism: data, tensor, pipeline, ZeRO – 70B Model on 64 GPUs #LLM #Training

L-37: Parallelism: data, tensor, pipeline, ZeRO (Hard) This video addresses the distributed training challenge of fitting a 70 billion ...

3:47

L-37: Parallelism: data, tensor, pipeline, ZeRO – 70B Model on 64 GPUs #LLM #Training

0 views

21 min ago

Morena

Somos cada vez más las personas usamos inteligencia artificial para escribir un mail, hacer un trabajo, o "mejorar" un párrafo.

30:12

¿Por qué hablamos como ChatGPT?

1 view

47 min ago

Mehdi Hosseini Moghadam

L-40: Greedy, top-k, top-p – LLM Decoding: Sampling #llm #decoding #python

L-40: Greedy, top-k, top-p (Easy) This video covers the core concept of LLM decoding, which converts raw model logits into ...

4:06

L-40: Greedy, top-k, top-p – LLM Decoding: Sampling #llm #decoding #python

0 views

19 min ago

Mehdi Hosseini Moghadam

L-29: Build a GPT end to end – Transformer Blocks to Text Generation #gpt #transformer #python

L-29: Build a GPT end to end (Hard) This video addresses the challenge of constructing a complete Generative Pre-trained ...

4:32

L-29: Build a GPT end to end – Transformer Blocks to Text Generation #gpt #transformer #python

0 views

29 min ago

Dennis Mohle

Synthetic Dissonance The Psychology of AI Gaslighting

Created by Gemini Notebook from these two sources: FELIX AND THE FAILSAFE FALLACY: NAVIGATING CONTEXTUAL ...

4:11

Synthetic Dissonance The Psychology of AI Gaslighting

0 views

14 min ago

DeReticular

Bounding Autonomous Swarms Replacing Perimeter SSO with Continuous Multi-Agent Execution Telemetry

The provided sources outline the **Resilient Epistemic & Thermodynamic Ledger Architecture (RELA)** and **Durable ...

10:12

Bounding Autonomous Swarms Replacing Perimeter SSO with Continuous Multi-Agent Execution Telemetry

1 view

1h ago

AI Profit Boardroom

Cut your AI token costs with this simple trick.

Want to make money and save time with AI? Join here: https://www.skool.com/ai-profit-lab-7462/about Stop wasting Claude and ...

1:01

Cut your AI token costs with this simple trick.

0 views

6 min ago

ViewTube