paper.dinq.meAI 与前沿科学顶会 · 学术新星雷达
登录 DINQ
返回新星榜
NS

Nikita Sorokin

SIGIR 2026×2代码可解释性效率与系统强化学习与后训练评测与基准
新星分新星
78
一作贡献70
核心/参与0
方向广度10
合作网络3
2 一作 · 2 篇 · SIGIR 2026

获取完整研究档案

首次获取该候选人的完整档案及 StartLux 评估共 20 积分;每位用户单独解锁,之后查看免费。主动刷新 20 积分。

联系方式

实时查找电子邮箱,费用不超过20积分;未找到邮箱不扣费。

2026 顶会论文

2
一作SIGIR 2026Oral· Information Retrieval (Full)

Hierarchical Embedding Fusion for Retrieval-Augmented Code Generation

Nikita Sorokin · Ivan Sedykh · Valentin Malykh

摘要

Retrieval-augmented code generation commonly conditions a decoder on large retrieved snippets, which couples online cost to repository size and introduces long-context noise. We present Hierarchical Embedding Fusion (HEF), a two-stage repository representation for code completion: (i) an offline...

可解释性效率与系统代码OpenReview
一作SIGIR 2026Poster· Information Retrieval (Short)

SAGE: Solver-Aligned Guided Exploration

Nikita Sorokin · Ivan Sedykh · Timur Ionov · Valentin Malykh

摘要

Modern code agents achieve strong results on challenging software engineering benchmarks such as SWE-bench, but solving each issue remains expensive: most inference cost is spent not on patch generation, but on repository exploration and search, accounting for up to 56% of tokens in our experimen...

代码强化学习与后训练评测与基准OpenReview

常见合作者