Post-Training Research Scientist - Forecasting and Reasoning

Point72 • New York, NY

2d

About The Position

Point72 develops and uses large language models (LLMs) for forecasting and decision-making in financial markets. Its research combines structured time series, unstructured information, and computational tools to study how models acquire evidence, make calibrated predictions, and revise beliefs as new information arrives. The work uses proprietary datasets, domain expertise, and real investment applications to evaluate model capabilities with rich, challenging, noisy, and non-stationary data. Research focuses on training environments, reward design, policy optimization, post-training, and evaluation.

The Post-Training Research Scientist will develop methodologies spanning environment and state construction, data generation and mixing, training, and evaluation. Working with other researchers, the scientist will help shape the research agenda and translate successful methods into production. The role involves developing post-training methods for forecasting and reasoning; constructing point-in-time environments and task distributions; developing reward modeling and credit assignment; collaborating on efficient training systems for multi-node GPU clusters; establishing rigorous evaluations; and productionizing research in collaboration with quantitative researchers.

Requirements

• Bachelor’s degree or higher in a quantitative or technical discipline, or equivalent practical experience developing advanced learning systems.
• Demonstrated expertise in large language model post-training, reinforcement learning, environment design, or preference learning.
• Hands-on experience with distributed GPU training, including performance optimization, numerical stability, checkpointing, and reproducibility.
• Strong implementation skills in Python and PyTorch or JAX for research and production model training.
• Experience using distributed training and inference frameworks, including FSDP, DeepSpeed, SGLang, Megatron, verl, or Slime.
• Ability to design rigorous evaluations for forecasting accuracy, calibration, reasoning quality, robustness, temporal holdouts, and leakage detection.
• Commitment to the highest ethical standards.

Nice To Haves

• Experience working in non-fully verifiable domains.
• Experience with probabilistic forecasting, time-series modeling, asynchronous reinforcement learning, tool-using agents, or production post-training systems, whether in academic research, frontier AI labs, or teams developing advanced learning systems.

Responsibilities

• Develop post-training methods for forecasting and reasoning, including supervised fine-tuning, reinforcement learning, on-policy distillation, and related methods to improve accuracy and calibration.
• Build point-in-time environments combining historical observations, documents, prior predictions, and tools, and define tasks involving information gathering, belief updating, and forecasting across multiple horizons.
• Develop reward modeling and credit assignment methods that combine verifiable intermediate outcomes, expert supervision, preference data, and delayed feedback.
• Collaborate with engineers to develop synchronous and asynchronous learning pipelines on multi-node GPU clusters.
• Establish rigorous evaluations of forecasting accuracy, uncertainty calibration, reasoning quality, robustness, and downstream utility, including temporal holdouts and controlled ablations to distinguish generalization from leakage, distribution shifts, and evaluator noise.
• Collaborate with quantitative researchers to identify valuable model capabilities and deploy successful methods into reliable production systems.

Benefits

• Fully paid health care benefits.
• Generous parental and family leave policies.
• Volunteer opportunities.
• Support for employee-led affinity groups representing women, people of color, and the LGBT+ community.
• Mental and physical wellness programs.
• Tuition assistance.
• A 401(k) savings program with an employer match.

What This Job Offers

• Job Type: Full-time
• Career Level: Senior