Daman Arora

Language models can be trained to proactively avoid costly errors by learning to "fail gracefully" directly from outcome-level rewards, without needing explicit safety annotations or verifiers.

Modifying executable agent harnesses rather than model weights allows test-time adaptation to generalize across unseen reasoning tasks without a single gradient step.

Direct thread communication in MPLMs cuts context requirements by half, revolutionizing how LLMs tackle complex reasoning tasks.