6 papers · 1 filter
Step Rejection Fine-Tuning: A Practical Distillation Recipe
Igor Slinko, Ilia Zavidnyi, Egor Bogomolov +1
Rejection Fine-Tuning (RFT) is a standard method for training LLM agents, where unsuccessful trajectories are discarded from the training set. In the context of SWE-bench tasks, th…
The Best of N Worlds: Aligning Reinforcement Learning with Best-of-N Sampling via max@k Optimisation
Farid Bagirov, Mikhail Arkhipov, Ksenia Sycheva +2
The application of Reinforcement Learning with Verifiable Rewards (RLVR) to mathematical and coding domains has demonstrated significant improvements in the reasoning and problem-s…
Practical Code RAG at Scale: Task-Aware Retrieval Design Choices under Compute Budgets
Timur Galimzyanov, Olga Kolomyttseva, Egor Bogomolov
We study retrieval design for code-focused generation tasks under realistic compute budgets. Using two complementary tasks from Long Code Arena -- code completion and bug localizat…
PIPer: On-Device Environment Setup via Online Reinforcement Learning
Alexander Kovrigin, Aleksandra Eliseeva, Konstantin Grotov +2
Environment setup-the process of configuring the system to work with a specific software project-represents a persistent challenge in Software Engineering (SE). Automated environme…
EnvBench: A Benchmark for Automated Environment Setup
Aleksandra Eliseeva, Alexander Kovrigin, Ilia Kholkin +2
Recent advances in Large Language Models (LLMs) have enabled researchers to focus on practical repository-level tasks in software engineering domain. In this work, we consider a co…
Long Code Arena: a Set of Benchmarks for Long-Context Code Models
Egor Bogomolov, Aleksandra Eliseeva, Timur Galimzyanov +8
Nowadays, the fields of code and natural language processing are evolving rapidly. In particular, models become better at processing long context windows - supported context sizes…