Co-Evolution of Policy and Internal Reward for Language Agents

1. 1Introduce self-generated internal rewards to language agents
2. 2Use reinforcement learning to co-evolve policy and internal reward
3. 3Evaluate the performance of language agents using sparse and delayed rewards
4. 4Analyze the impact of self-generated internal rewards on policy improvement

ML researchers and AI engineers can benefit from this approach to improve language agent training, as it allows for more efficient and effective learning

💡 Self-generated internal rewards can improve policy and reward co-evolution in language agents

💡 Self-Guide: self-generated internal rewards for language agents