Remote Multimodal RL Research Engineer

Winclap; United States; Remoto; USD 120.000 - 180.000; Tempo pieno

Descrizione del lavoro

Tether in the United States seeks a senior researcher to design and optimize multi-modal reinforcement learning systems. You will work with diffusion-based and autoregressive models across text, image, and audio, building scalable RL pipelines, simulation environments, and evaluation frameworks to drive robust, domain-specific AI performance.

Strong candidates bring deep RL theory, hands-on PyTorch experience, and a proven publication record in top conferences.

Competenze

• Requires a Master's degree in Computer Science or a related field; a PhD in ML/NLP/CV is preferred.
• Demonstrated experience running large-scale reinforcement learning experiments in multimodal/vision-centric systems.
• Deep understanding of RL algorithms and optimization for vision and multimodal problems.
• Strong proficiency in PyTorch and deep learning frameworks; experience building RL pipelines.
• Empirical research focus with ability to improve RL challenges in multimodal/vision tasks.
• Proven track record of publications in top-tier conferences (ICML, NeurIPS, ICLR, CVPR, ICCV, ECCV).

Mansioni

• Conduct research on reinforcement learning algorithms for multimodal models and unified frameworks.
• Design and build scalable RL infrastructure for distributed training across multimodal systems.
• Develop reward modeling strategies to improve training stability and mitigate failures.
• Create and curate multimodal simulation environments and datasets for robust training and benchmarking.
• Design rigorous benchmarking protocols to evaluate model performance across modalities.
• Analyze and optimize policy performance across modalities, addressing bottlenecks in training and cross-modal alignment.
• Explore next-gen RL paradigms to achieve superior domain-adapted AI performance.
• Publish research findings in top conferences.

Formazione

Master's degree in Computer Science or related field

PhD in Machine Learning / NLP / Computer Vision (preferred)

Reinforcement learning

Multimodal learning

Policy stability / exploration / sample efficiency

Research publications (top-tier conferences)

Strumenti

PyTorch

End-to-end RL pipelines

Vision and multimodal AI frameworks

Descrizione del lavoro

Tether in the United States seeks a senior researcher to design and optimize multi-modal reinforcement learning systems. You will work with diffusion-based and autoregressive models across text, image, and audio, building scalable RL pipelines, simulation environments, and evaluation frameworks to drive robust, domain-specific AI performance.

Strong candidates bring deep RL theory, hands-on PyTorch experience, and a proven publication record in top conferences.