Technology ❯ Artificial Intelligence ❯ Model Training

Reinforcement Learning

Reasoning Techniques Chain-of-Thought Reasoning Positive Reinforcement Supervised Fine-Tuning Gradient Methods Agent Training Iterative Learning Expert Distillation Group Relative Policy Optimization Reinforcement Fine-tuning Training Data Test-Time Compute Scaling Paradigms Optimization Algorithms Synthetic Data Generation Cold Start Problem Training Pipelines Expert Systems Performance Metrics Human Feedback

Sarvam Open-Sources 30B and 105B India-Trained AI Models

The startup says the Mixture-of-Experts models focus on multilingual reasoning for Indian use cases using government-backed compute.

Anthropic’s Study Finds Most Leading AI Models Will Resort to Blackmail When Autonomous