Post-Training Research Scientist - Forecasting and Reasoning

Point72
New York, NY
On-site

Who this role is best for

Strong fit for researchers with post-training LLM expertise who build environments and design reward systems, working in financial markets with a focus on production deployment and rigorous evaluation.

Best fit for

  • Candidates with post-training LLM experience and a background in financial modeling or trading systems
    — “develop and use LLMs for forecasting and decision-making in financial markets”
  • Individuals skilled in distributed GPU training and production pipeline development
    — “Build efficient, reliable training systems. Collaborate with engineers to develop synchronous and asynchronous learning pipelines on multi-node GPU clusters.”
  • Researchers who can translate complex methodologies into real-world investment applications
    — “translate successful methods into production”

Things to consider

  • The role demands a high level of calibration and evaluation rigor for forecasting and reasoning tasks
    — “Establish rigorous evaluations that measure forecasting accuracy, uncertainty calibration, reasoning quality, robustness, and downstream utility.”
  • Experience with non-fully verifiable domains is preferred, indicating potential ambiguity in validation processes
    — “Experience working in non-fully verifiable domains is preferred.”

How to stand out

  • Highlight experience with temporal holdouts and controlled ablations in model evaluations
    — “Design temporal holdouts and controlled ablations that distinguish generalization from leakage”
  • Emphasize contributions to production systems, not just research outputs
    — “Productionize research. Collaborate with quantitative researchers to identify valuable model capabilities”
  • Showcase work with reinforcement learning and belief updating in dynamic environments
    — “Define tasks that require information gathering, belief updating, and forecasting across multiple horizons”
Pace · Fast PacedCollaboration · HighAutonomy · MediumDecision Impact · Company

Derived from job-description analysis by Serendipath's career intelligence engine.

What success looks like

  • accurate forecasting
  • calibrated predictions
  • robust models
Typical background
quantitative researchAI research

Skills & requirements

Required

Large Language Model Post-trainingReinforcement LearningEnvironment DesignReward ModelingDistributed GPU Training

Preferred

Probabilistic ForecastingTime-series ModelingAsynchronous Reinforcement Learning

Stack & domain

Large Language Model (llm) Post-trainingReinforcement LearningEnvironment DesignPreference LearningDistributed GPU TrainingPythonPyTorchJAXFSDPDeepspeedSglangMegatronVerlSlimeTeamworkProblem-solvingCommunicationAttention To DetailEthicsFinancial MarketsForecastingReasoning

About the role

Original posting from Point72 via Greenhouse

POST-TRAINING RESEARCH SCIENTIST - FORECASTING AND REASONING 

We develop and use LLMs for forecasting and decision-making in financial markets. Our research combines structured time series, unstructured information, and computational tools to study how models acquire useful evidence, make calibrated predictions, and revise their beliefs as new information arrives. We rely on proprietary datasets, domain expertise, and real investment applications to evaluate these capabilities. We deal with rich, challenging, noisy, non-stationary data. Our work focuses on advances in training environments, reward design, policy optimization, post-training and evaluation.

WHAT YOU'LL DO

We are looking for an exceptional post-training research scientist to develop methodologies across environment and state construction, data generation and mixing, training, and evaluation. Working closely with other researchers, you will help shape the research agenda and translate successful methods into production. Specifically, you will:

Develop post-training methods for forecasting and reasoning. Design and execute supervised fine-tuning, reinforcement learning, on-policy distillation, and related methods to improve accuracy and calibration.

Build environments and task distributions. Construct point-in-time environments combining historical observations, documents, prior predictions, and tools. Define tasks that require information gathering, belief updating, and forecasting across multiple horizons.

Develop reward modeling and credit assignment methods, combining verifiable intermediate outcomes, expert supervision, preference data, and delayed feedback.

Build efficient, reliable training systems. Collaborate with engineers to develop synchronous and asynchronous learning pipelines on multi-node GPU clusters.

Establish rigorous evaluations that measure forecasting accuracy, uncertainty calibration, reasoning quality, robustness, and downstream utility. Design temporal holdouts and controlled ablations that distinguish generalization from leakage, distribution shifts, and evaluator noise.

Productionize research. Collaborate with quantitative researchers to identify valuable model capabilities and deploy successful methods into reliable production systems.

WHAT'S REQUIRED

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 (LLM) post-training, reinforcement learning, environment design, or preference learning. Experience working in non-fully verifiable domains is preferred.

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.

Experience with probabilistic forecasting, time-series modeling, asynchronous reinforcement learning (RL), tool-using agents, or production post-training systems from academic research, frontier AI labs, or teams developing advanced learning systems.

Commitment to the highest ethical standards.

WE TAKE CARE OF OUR PEOPLE

We invest in our people, their careers, their health, and their well-being. When you work here, we provide:

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 and more

ABOUT POINT72

Point72 is a leading global alternative investment firm led by Steven A. Cohen. Building on more than 30 years of investing experience, Point72 seeks to deliver superior returns for its investors through fundamental and systematic investing strategies across asset classes and geographies. We aim to attract and retain the industry's brightest talent by cultivating an investor-led culture and committing to our people's long-term growth. For more information, visit https://point72.com/.

Source: Point72 careers (Greenhouse)

Similar roles

More roles at Point72