AI.info

Rohit Thekkanal

Explore Rohit Thekkanal on AI.info.

• Breaking the Safety-Capability Tradeoff: Reinforcement Learning with Verifiable Rewards Maintains Safety Guardrails in LLMs