Archives des Laboratory for Information and Decision Systems (LIDS)

Study: Platforms that rank the latest LLMs can be unreliable
16 Fév
Why it’s critical to move beyond overly aggregated machine-learning metrics
16 Fév
3 Questions: How AI could optimize the power grid
16 Fév
MIT scientists investigate memorization risk in the age of clinical AI
16 Fév
New method improves the reliability of statistical estimations
16 Fév
A smarter way for large language models to think about hard problems
16 Fév
MIT engineers design an aerial microrobot that can fly as fast as a bumblebee
16 Fév
New control system teaches soft robots the art of staying safe
16 Fév
Researchers discover a shortcoming that makes LLMs less reliable
16 Fév
Charting the future of AI, from safer answers to faster thinking
16 Fév
Teaching robots to map large environments
16 Fév
A faster problem-solving tool that guarantees feasibility
16 Fév
Fighting for the health of the planet with AI
16 Fév
New prediction model could improve the reliability of fusion power plants
16 Fév
3 Questions: The pros and cons of synthetic data in AI
16 Fév
A new way to test how well AI systems classify text
16 Fév
Eco-driving measures could significantly reduce vehicle emissions
16 Fév
New algorithms enable efficient machine learning with symmetric data
16 Fév
A new way to edit or generate images
16 Fév
This “smart coach” helps LLMs switch between text and code
16 Fév
How to more efficiently study complex treatment interactions
16 Fév
Confronting the AI/energy conundrum
16 Fév
LLMs factor in unrelated information when recommending medical treatments
16 Fév
Unpacking the bias of large language models
16 Fév
Inroads to personalized AI trip planning
16 Fév
Melding data, systems, and society
16 Fév
AI-enabled control system helps autonomous drones stay on target in uncertain environments
16 Fév
An anomaly detection framework anyone can use
16 Fév
The sweet taste of a new idea
16 Fév
With AI, researchers predict the location of virtually any protein within a human cell
16 Fév
Study shows vision-language models can’t handle queries with negation words
16 Fév
New tool evaluates progress in reinforcement learning
16 Fév
Designing a new way to optimize complex coordinated systems
16 Fév
A faster way to solve complex planning problems
16 Fév
Researchers teach LLMs to solve complex planning challenges
16 Fév

Tag

#Laboratory for Information and Decision Systems (LIDS)

Study: Platforms that rank the latest LLMs can be unreliable

Why it’s critical to move beyond overly aggregated machine-learning metrics

3 Questions: How AI could optimize the power grid

MIT scientists investigate memorization risk in the age of clinical AI

New method improves the reliability of statistical estimations

A smarter way for large language models to think about hard problems

MIT engineers design an aerial microrobot that can fly as fast as a bumblebee

New control system teaches soft robots the art of staying safe

Researchers discover a shortcoming that makes LLMs less reliable

Charting the future of AI, from safer answers to faster thinking

Teaching robots to map large environments

A faster problem-solving tool that guarantees feasibility

Fighting for the health of the planet with AI

New prediction model could improve the reliability of fusion power plants

3 Questions: The pros and cons of synthetic data in AI

A new way to test how well AI systems classify text

Eco-driving measures could significantly reduce vehicle emissions

New algorithms enable efficient machine learning with symmetric data

A new way to edit or generate images

This “smart coach” helps LLMs switch between text and code

How to more efficiently study complex treatment interactions

Confronting the AI/energy conundrum

LLMs factor in unrelated information when recommending medical treatments

Unpacking the bias of large language models

Inroads to personalized AI trip planning

Melding data, systems, and society

AI-enabled control system helps autonomous drones stay on target in uncertain environments

An anomaly detection framework anyone can use

With AI, researchers predict the location of virtually any protein within a human cell

Study shows vision-language models can’t handle queries with negation words

New tool evaluates progress in reinforcement learning

Designing a new way to optimize complex coordinated systems

A faster way to solve complex planning problems

Researchers teach LLMs to solve complex planning challenges

Study: Platforms that rank the latest LLMs can be unreliable

Why it’s critical to move beyond overly aggregated machine-learning metrics

3 Questions: How AI could optimize the power grid

MIT scientists investigate memorization risk in the age of clinical AI

New method improves the reliability of statistical estimations

A smarter way for large language models to think about hard problems

MIT engineers design an aerial microrobot that can fly as fast as a bumblebee

New control system teaches soft robots the art of staying safe

Researchers discover a shortcoming that makes LLMs less reliable

Charting the future of AI, from safer answers to faster thinking

Teaching robots to map large environments

A faster problem-solving tool that guarantees feasibility

Fighting for the health of the planet with AI

New prediction model could improve the reliability of fusion power plants

3 Questions: The pros and cons of synthetic data in AI

A new way to test how well AI systems classify text

Eco-driving measures could significantly reduce vehicle emissions

New algorithms enable efficient machine learning with symmetric data

A new way to edit or generate images

This “smart coach” helps LLMs switch between text and code

How to more efficiently study complex treatment interactions

Confronting the AI/energy conundrum

LLMs factor in unrelated information when recommending medical treatments

Unpacking the bias of large language models

Inroads to personalized AI trip planning

Melding data, systems, and society

AI-enabled control system helps autonomous drones stay on target in uncertain environments

An anomaly detection framework anyone can use

The sweet taste of a new idea

With AI, researchers predict the location of virtually any protein within a human cell

Study shows vision-language models can’t handle queries with negation words

New tool evaluates progress in reinforcement learning

Designing a new way to optimize complex coordinated systems

A faster way to solve complex planning problems

Researchers teach LLMs to solve complex planning challenges