- agents
- multimodal-ai
- ai-salon
- research
•
•
•
-
Level up your Agents: Teaching Vision-Language Models to Play by the Rules
We explore how Vision-Language Models can be improved for interactive decision-making by using our new reinforcement learning technique called Advantage-Filtered Supervised Fine-Tuning.
-
Will AI Empower or Eclipse Human Creativity? An Emerging Question in the Age of Intelligent Machines
A debate of whether AI will empower or eclipse human creativity, considering its potential as a powerful tool versus the risk of over-reliance diminishing our innate abilities.
-
The Generalization Gambit - Does It Still Matter in the Age of Big AI?
Big AI models are now trained on tons of data, is "generalization" (performing on new data) still important? Is generalization an outdated concept or crucial for real-world reliability?
-
Are your Visual Programs Right for the Wrong Reasons?
ViUniT, Making AI visual reasoning reliable, one unit test at a time.
-
Introducing TACO - Salesforce AI Research's Family of Multimodal Action Models - Salesforce
We present TACO, a family of multi-modal large action models designed to improve performance on complex questions that require multiple capabilities and demand multi-step solutions.