Research Associate - LTI - School of Computer Science

Carnegie Mellon University • Pittsburgh, PA

46d

About The Position

Carnegie Mellon University is a private, global research university that stands among the world’s most renowned education institutions. With ground-breaking brain science, path-breaking performances, creative start-ups, big data, big ambitions, hands-on learning, and a whole lot of robots, CMU doesn’t imagine the future, we invent it. If you’re passionate about joining a community that challenges the curious to deliver work that matters, your journey starts here! The Language Technologies Institute (LTI) at CMU is a world-renowned research and education hub at the forefront of natural language processing, machine learning, artificial intelligence, and human-computer interaction. Established within the School of Computer Science, LTI pioneers innovative ways to understanding, processing, and generating human language in both written and spoken forms. We are seeking a motivated Research Associate to join our team to work on cutting-edge "Reasoning Reward Models" (RRMs). The purpose of this role is to help move beyond traditional scalar reward signals by developing models that generate natural language insights to explain human preferences. The intern will assist in implementing pipelines that bridge structured logical reasoning (e.g., Math) and open-ended human reasoning (e.g., Medical), ultimately contributing to new methods for Large Language Model (LLM) post-training and alignment.

Requirements

• Bachelor's Degree in Computer Science, AI, ML, Data Science or a related field.
• Programming proficiency in Python, with experience in deep learning.
• Knowledge and experience of LLM fundamentals
• Academic or project based familiarly with Large Language Model Concepts, including Supervised Fine-Tuning (SFT) and basic Reinforcement Learning (RL) concepts
• Experience manipulating and processing datasets for NLP tasks (working with tokenizers, JSONL formats or Hugging Face datasets).
• Demonstrated ability to read technical research papers and implement algorithms or baselines from code repositories.
• A combination of education and meaningful experience from which comparable knowledge is proven may be considered.
• Successful background check investigation may be required

Responsibilities

• Train Reasoning Reward Models using Supervised Fine-Tuning (SFT) and Direct Preference Optimization (DPO) to generate natural-language insights for Math and Medical tasks.
• Implement insight-based post-training mechanisms, specifically "Behavioral Priming" and "Insight Exhibition Rewards," to align models with interpretable reasoning criteria.
• Develop and implement software prototypes, algorithms, and data pipelines to support ongoing research projects.
• Execute experimental workflows and simulations, ensuring accurate data collection and logging.
• Analyze experimental results and metrics to identify trends, errors, or areas for optimization.
• Document technical processes, codebases, and research findings to ensure reproducibility and knowledge transfer.
• Maintain up-to-date knowledge of relevant tools, libraries, and best practices in software engineering and research.

Benefits

• Benefits eligible employees enjoy a wide array of benefits including comprehensive medical, prescription, dental, and vision insurance as well as a generous retirement savings program with employer contributions.
• Unlock your potential with tuition benefits
• take well-deserved breaks with ample paid time off and observed holidays
• rest easy with life and accidental death and disability insurance.
• Additional perks include a free Pittsburgh Regional Transit bus pass, access to our Family Concierge Team to help navigate childcare needs, fitness center access , and much more!

What This Job Offers

• Job Type: Full-time
• Career Level: Entry Level
• Number of Employees: 5,001-10,000 employees