◍wovepaper
SearchResearchersInstitutions
Sign in
researcher

B. Shaposhnikov

8 papers hereh-index 592 citations8 works total

Matching runs newest-first, so older work may not be attached to this profile yet.

author position
  • middle author8

Across the 8 of 8 papers where every author was matched, so the position is known.

fields
  • cs.LG8

identity via Semantic Scholar / OpenAlex

activity
20242026
collaborators
Showing 2025 · cs.LGShow all

4 papers · 2 filters

cs.LG2025

Small Vectors, Big Effects: A Mechanistic Study of RL-Induced Reasoning via Steering Vectors

Viacheslav Sinii, Nikita Balagansky, Gleb Gerasimov +6

The mechanisms by which reasoning training reshapes LLMs' internal computations remain unclear. We study lightweight steering vectors inserted into the base model's residual stream…

cs.LG2025

ESSA: Evolutionary Strategies for Scalable Alignment

Daria Korotyshova, Boris Shaposhnikov, Alexey Malakhov +7

Online alignment of large language models (LLMs) is dominated by reinforcement learning from human feedback (RLHF) with gradient-based optimizers such as PPO or GRPO. While effecti…

cs.LG2025

Steering LLM Reasoning Through Bias-Only Adaptation

Viacheslav Sinii, Alexey Gorbatovski, Artem Cherepanov +3

We show that training a single d-dimensional steering vector per layer with reinforcement learning, while freezing all base weights, matches the accuracy of fully RL-tuned reason…

cs.LG2025

The Differences Between Direct Alignment Algorithms are a Blur

Alexey Gorbatovski, Boris Shaposhnikov, Viacheslav Sinii +2

Direct Alignment Algorithms (DAAs) simplify LLM alignment by directly optimizing policies, bypassing reward modeling and RL. While DAAs differ in their use of SFT (one-stage vs. tw…

◍wovepaper

Papers, researchers and institutions, woven together.

Explore
  • Search
  • Researchers
  • Institutions
Account
  • Library
  • Chat
Data
  • arXiv.org
  • Semantic Scholar
  • OpenAlex
  • Latest RSS
AboutContactPrivacyDevelopersllms.txtopenapi.json
Not affiliated with arXiv. Researcher data from Semantic Scholar (ODC-BY) and OpenAlex.