Serguei Barannikov — accepted AI-conference papers

2025

Feature-Level Insights into Artificial Text Detection with Sparse Autoencoders

Kristian Kuznetsov, Laida Kushnareva, Anton Razzhigaev, Polina Druzhinina, Anastasia Voznyuk, Irina Piontkovskaya +2

ACL 2025finding

Artificial Text Detection (ATD) is becoming increasingly important with the rise of advanced Large Language Models (LLMs). Despite numerous efforts, no single algorithm performs consistently well across different types of unseen text or guarantees effective generalization to new LLMs. Interpretabili…

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗

2025

Quantifying Logical Consistency in Transformers via Query-Key Alignment

Eduard Tulchinskii, Laida Kushnareva, Anastasia Voznyuk, Andrei Andriiainen, Irina Piontkovskaya, Evgeny Burnaev +1

EMNLP 2025

Large language models (LLMs) excel at many NLP tasks, yet their multi-step logical reasoning remains unreliable. Existing solutions such as Chain-of-Thought prompting generate intermediate steps but provide no internal check of their logical coherence. In this paper, we use the “QK-score”, a lightwe

Cited by 0 Source ↗PDF ↗Scholar ↗

2025

RTD-Lite: Scalable Topological Analysis for Comparing Weighted Graphs in Learning Tasks

Eduard Tulchinskii, Daria Voronkova, Ilya Trofimov, Evgeny Burnaev, Serguei Barannikov

AISTATS 2025poster

Topological methods for comparing weighted graphs are valuable in various learning tasks but often suffer from computational inefficiency on large datasets. We introduce RTD-Lite, a scalable algorithm that efficiently compares topological features, specifically connectivity or cluster structures at…

Cited by 0 Source ↗code ↗Scholar ↗

2024

Disentanglement Learning via Topology

Nikita Balabin, Daria Voronkova, Ilya Trofimov, Evgeny Burnaev, Serguei Barannikov

ICML 2024poster

We propose TopDis (Topological Disentanglement), a method for learning disentangled representations via adding a multi-scale topological loss term. Disentanglement is a crucial property of data representations substantial for the explainability and robustness of deep learning models and a step towar…

Cited by 7 Source ↗PDF ↗code ↗Scholar ↗

2024

Robust AI-Generated Text Detection by Restricted Embeddings

Kristian Kuznetsov, Eduard Tulchinskii, Laida Kushnareva, German Magai, Serguei Barannikov, Sergey Nikolenko +1

EMNLP 2024finding

Growing amount and quality of AI-generated texts makes detecting such content more difficult. In most real-world scenarios, the domain (style and topic) of generated data and the generator model are not known in advance. In this work, we focus on the robustness of classifier-based detectors of AI-ge…

Cited by 4 Source ↗PDF ↗code ↗Scholar ↗

2024

Scalar Function Topology Divergence: Comparing Topology of 3D Objects

Ilya Trofimov*, Daria Voronkova, Eduard Tulchinskii, Evgeny Burnaev, Serguei Barannikov

ECCV 2024poster

"We propose a new topological tool for computer vision - Scalar Function Topology Divergence (SFTD), which measures the dissimilarity of multi-scale topology between sublevel sets of two functions having a common domain. Functions can be defined on an undirected graph or Euclidean space of any dimen…

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗

2023

Intrinsic Dimension Estimation for Robust Detection of AI-Generated Texts

Eduard Tulchinskii, Kristian Kuznetsov, Kushnareva Laida, Daniil Cherniavskii, Sergey Nikolenko, Evgeny Burnaev +2

NeurIPS 2023poster

Rapidly increasing quality of AI-generated content makes it difficult to distinguish between human and AI-generated texts, which may lead to undesirable consequences for society. Therefore, it becomes increasingly important to study the properties of human texts that are invariant over text domains…

Cited by 82 Source ↗PDF ↗code ↗Scholar ↗

2023

Learning topology-preserving data representations

Ilya Trofimov, Daniil Cherniavskii, Eduard Tulchinskii, Nikita Balabin, Evgeny Burnaev, Serguei Barannikov

ICLR 2023poster

We propose a method for learning topology-preserving data representations (dimensionality reduction). The method aims to provide topological similarity between the data manifold and its latent representation via enforcing the similarity in topological features (clusters, loops, 2D voids, etc.) and…

Cited by 26 Source ↗PDF ↗code ↗Scholar ↗

2022

Acceptability Judgements via Examining the Topology of Attention Maps

Daniil Cherniavskii, Eduard Tulchinskii, Vladislav Mikhailov, Irina Proskurina, Laida Kushnareva, Ekaterina Artemova +4

EMNLP 2022finding

The role of the attention mechanism in encoding linguistic knowledge has received special interest in NLP. However, the ability of the attention heads to judge the grammatical acceptability of a sentence has been underexplored. This paper approaches the paradigm of acceptability judgments with topol…

Cited by 24 Source ↗PDF ↗code ↗Scholar ↗

2022

Representation Topology Divergence: A Method for Comparing Neural Network Representations.

Serguei Barannikov, Ilya Trofimov, Nikita Balabin, Evgeny Burnaev

ICML 2022spotlight

Comparison of data representations is a complex multi-aspect problem. We propose a method for comparing two data representations. We introduce the Representation Topology Divergence (RTD) score measuring the dissimilarity in multi-scale topology between two point clouds of equal size with a one-to-o…

Cited by 70 Source ↗PDF ↗code ↗Scholar ↗

2021

Artificial Text Detection via Examining the Topology of Attention Maps

Laida Kushnareva, Daniil Cherniavskii, Vladislav Mikhailov, Ekaterina Artemova, Serguei Barannikov, Alexander Bernstein +3

EMNLP 2021main

The impressive capabilities of recent generative models to create texts that are challenging to distinguish from the human-written ones can be misused for generating fake news, product reviews, and even abusive content. Despite the prominent performance of existing methods for artificial text detect…

Cited by 56 Source ↗PDF ↗code ↗Scholar ↗

2021

Manifold Topology Divergence: a Framework for Comparing Data Manifolds.

Serguei Barannikov, Ilya Trofimov, Grigorii Sotnikov, Ekaterina Trimbach, Alexander Korotin, Alexander Filippov +1

NeurIPS 2021poster

We propose a framework for comparing data manifolds, aimed, in particular, towards the evaluation of deep generative models. We describe a novel tool, Cross-Barcode(P,Q), that, given a pair of distributions in a high-dimensional space, tracks multiscale topology spacial discrepancies between manifol…

Cited by 35 Source ↗PDF ↗code ↗Scholar ↗