7 papers
Predictable GRPO: A Closed-Form Model of Training Dynamics
Rajat Ghosh, Datta Nimmaturi, Aryan Singhal +4
We develop a first-principles reduced-order model of these dynamics. Under a single mean-field assumption that summarizes the policy by its expected reward, we reduce the GRPO upda…
Go-UT-Bench: A Fine-Tuning Dataset for LLM-Based Unit Test Generation in Go
Yashshi Pipalani, Hritik Raj, Rajat Ghosh +2
Training data imbalance poses a major challenge for code LLMs. Most available data heavily over represents raw opensource code while underrepresenting broader software engineering…
Predictive Scaling Laws for Efficient GRPO Training of Large Reasoning Models
Datta Nimmaturi, Vaishnavi Bhargava, Rajat Ghosh +2
Fine-tuning large language models (LLMs) for reasoning tasks using reinforcement learning methods like Group Relative Policy Optimization (GRPO) is computationally expensive. To ad…
SWE-Tester: Training Open-Source LLMs for Issue Reproduction in Real-World Repositories
Aditya Bharat Soni, Rajat Ghosh, Vaishnavi Bhargava +2
Software testing is crucial for ensuring the correctness and reliability of software systems. Automated generation of issue reproduction tests from natural language issue descripti…
BAR Conjecture: the Feasibility of Inference Budget-Constrained LLM Services with Authenticity and Reasoning
Jinan Zhou, Rajat Ghosh, Vaishnavi Bhargava +2
When designing LLM services, practitioners care about three key properties: inference-time budget, factual authenticity, and reasoning capacity. However, our analysis shows that no…
The ALCHEmist: Automated Labeling 500x CHEaper Than LLM Data Annotators
Tzu-Heng Huang, Catherine Cao, Vaishnavi Bhargava +1
Large pretrained models can be used as annotators, helping replace or augment crowdworkers and enabling distilling generalist models into smaller specialist models. Unfortunately,…