DEV Community

Plain English Papers Series' Articles

Back to Mike Young's Series
Meet CulturaX: A New Multilingual Dataset for Training AI Models in 167 Languages
Cover image for Meet CulturaX: A New Multilingual Dataset for Training AI Models in 167 Languages

Meet CulturaX: A New Multilingual Dataset for Training AI Models in 167 Languages

5
Comments
5 min read
Contrastive Decoding: A New Technique for Boosting Reasoning in Large Language Models
Cover image for Contrastive Decoding: A New Technique for Boosting Reasoning in Large Language Models

Contrastive Decoding: A New Technique for Boosting Reasoning in Large Language Models

3
Comments
5 min read
New Research Says LLMs Are Surprisingly Good at Compressing... Images and Audio?
Cover image for New Research Says LLMs Are Surprisingly Good at Compressing... Images and Audio?

New Research Says LLMs Are Surprisingly Good at Compressing... Images and Audio?

6
Comments
4 min read
Bringing Still Pictures to Life with Neural Motion Textures
Cover image for Bringing Still Pictures to Life with Neural Motion Textures

Bringing Still Pictures to Life with Neural Motion Textures

4
Comments
4 min read
By Teaching AI to Make Pictures and Write, Scientists Improve Its Grasp of Vision and Language
Cover image for By Teaching AI to Make Pictures and Write, Scientists Improve Its Grasp of Vision and Language

By Teaching AI to Make Pictures and Write, Scientists Improve Its Grasp of Vision and Language

1
Comments
3 min read
LongLoRA: A New, More Efficient Way to Fine-Tune LLMs
Cover image for LongLoRA: A New, More Efficient Way to Fine-Tune LLMs

LongLoRA: A New, More Efficient Way to Fine-Tune LLMs

1
Comments
3 min read
Meet GPT4Tools: teaching existing LLMs how to use tools for visual tasks
Cover image for Meet GPT4Tools: teaching existing LLMs how to use tools for visual tasks

Meet GPT4Tools: teaching existing LLMs how to use tools for visual tasks

Comments
5 min read
UNC Researchers Present VideoDirectorGPT: Using AI to Generate Multi-Scene Videos from Text
Cover image for UNC Researchers Present VideoDirectorGPT: Using AI to Generate Multi-Scene Videos from Text

UNC Researchers Present VideoDirectorGPT: Using AI to Generate Multi-Scene Videos from Text

3
Comments
6 min read
Researchers discover explicit registers eliminate vision transformer attention spikes
Cover image for Researchers discover explicit registers eliminate vision transformer attention spikes

Researchers discover explicit registers eliminate vision transformer attention spikes

5
Comments
5 min read
Infinite Context Windows? LLMs for Streaming Applications with Attention Sinks
Cover image for Infinite Context Windows? LLMs for Streaming Applications with Attention Sinks

Infinite Context Windows? LLMs for Streaming Applications with Attention Sinks

3
Comments
5 min read
Tool-Integrated Reasoning: A New Approach for Math-Savvy LLMs
Cover image for Tool-Integrated Reasoning: A New Approach for Math-Savvy LLMs

Tool-Integrated Reasoning: A New Approach for Math-Savvy LLMs

4
Comments
5 min read
Enabling Language Models to Implicitly Learn Self-Improvement
Cover image for Enabling Language Models to Implicitly Learn Self-Improvement

Enabling Language Models to Implicitly Learn Self-Improvement

3
Comments
5 min read
Decoding Speech from Brain Waves - A Breakthrough in Brain-Computer Interfaces
Cover image for Decoding Speech from Brain Waves - A Breakthrough in Brain-Computer Interfaces

Decoding Speech from Brain Waves - A Breakthrough in Brain-Computer Interfaces

1
Comments
4 min read
Researchers: Low-Resource Languages Can Easily Jailbreak LLMs
Cover image for Researchers: Low-Resource Languages Can Easily Jailbreak LLMs

Researchers: Low-Resource Languages Can Easily Jailbreak LLMs

3
Comments
4 min read
There are many ways to design roundabouts. Can AI choose the best one?
Cover image for There are many ways to design roundabouts. Can AI choose the best one?

There are many ways to design roundabouts. Can AI choose the best one?

3
Comments
5 min read
Anomaly Detection in Multivariate Time Series with... Diffusion Models?
Cover image for Anomaly Detection in Multivariate Time Series with... Diffusion Models?

Anomaly Detection in Multivariate Time Series with... Diffusion Models?

1
Comments
8 min read
How Sora (actually) works

How Sora (actually) works

10
Comments 2
6 min read
Up to 17% of AI conference reviews now written by AI

Up to 17% of AI conference reviews now written by AI

Comments
7 min read
The death of creativity
Cover image for The death of creativity

The death of creativity

47
Comments 15
6 min read
LLMs are secretly good at regression calculations

LLMs are secretly good at regression calculations

4
Comments
9 min read
Microsoft’s Phi-3 model is cool tech, but local LLMs are useless
Cover image for Microsoft’s Phi-3 model is cool tech, but local LLMs are useless

Microsoft’s Phi-3 model is cool tech, but local LLMs are useless

Comments
6 min read
LLMs can speak in JPEG
Cover image for LLMs can speak in JPEG

LLMs can speak in JPEG

Comments
6 min read
🧠 Training on code improves LLM performance on non-coding tasks

🧠 Training on code improves LLM performance on non-coding tasks

2
Comments
6 min read
LLMs will lie forever
Cover image for LLMs will lie forever

LLMs will lie forever

20
Comments 7
4 min read