ET
Elena Tutubalina
cs.CLcs.AIcs.LGcs.ROcs.CV
On Valency
published · living versionsW_8pemzcn3·v1 · currentpublished
Context or Knowledge: How LLMs Attribute Their Own Answers
with Kirill Afendulev, Alexey Dontsov, Alla Chepurova, Anton Korznikov
1 version
W_5khfgb7z·v1published
Preprints & journals
18 papers in the corpus · 2025–2026From Stories to Features: Scaling Concept Vectors for AI Alignment & Interpretability2610.24556v1 · Dmitry Scherbakov, Oleg Chernikov, Victoria Khamash et al.2026 · 0 citationsarXivon Valency
Sparse Transformer: a Controlled Testbed for Mechanistic Interpretability2610.19907v1 · Anton Korznikov, Alexey Dontsov, Elena Tutubalina2026 · 0 citationsarXiv
Structural Chain-of-Thought: Reasoning with Executable Computational Graphs2610.15107v1 · Petr Senatorov, Karim Vafin, Elena Tutubalina et al.2026 · 0 citationsarXiv
LANTERN: Illuminating Hidden Mathematical Knowledge in Language Models2609.32264v1 · Pavel Tikhonov, Elena Tutubalina, Ivan Oseledets et al.2026 · 0 citationsarXiv
What Attention Recalls and Recurrence Controls in Hybrid Language Models2609.04434v1 · Kirill Afendulev, Alexey Dontsov, Elena Tutubalina et al.2026 · 0 citationsarXiv
The More Popular, The Harder to Forget: Adaptive Popularity for LLM Unlearning2608.14229v1 · Anna Borisiuk, Andrey Savchenko, Alexander Panchenko et al.2026 · 0 citationsarXiv
Does VLA Even Know the Basics? Measuring Commonsense and World Knowledge Retention in Vision-Language-Action Models2606.19297v1 · Nikita Kachaev, Andrey Moskalenko, Matvey Skripkin et al.2026 · 0 citationsarXiv
Harnessing non-adversarial robustness in large language models2605.29816v1 · Qinghua Zhou, Ellina Aleshina, Andrey Lovyagin et al.2026 · 0 citationsarXiv
Evolutionary Search for Automated Design of Uncertainty Quantification Methods2604.03473v1 · Mikhail Seleznyov, Daniil Korbut, Viktor Moskvoretskii et al.2026 · 0 citationsarXiv
Breaking the Chain: A Causal Analysis of LLM Faithfulness to Intermediate Structures2603.16475v2 · Oleg Somov, Mikhail Chaichuk, Gleb Ershov et al.2026 · 0 citationsarXiv
The Chronicles of RiDiC: Generating Datasets with Controlled Popularity Distribution for Long-form Factuality Evaluation2604.00019v1 · Pavel Braslavski, Dmitrii Iarosh, Nikita Sushko et al.2026 · 0 citationsarXiv
Leveraging LLM Parametric Knowledge for Fact Checking without Retrieval2603.05471v1 · Artem Vazhentsev, Maria Marina, Daniil Moskovskiy et al.2026 · 0 citationsarXiv
Anatomy of Unlearning: The Dual Impact of Fact Salience and Model Fine-Tuning2602.19612v5 · Anna Borisiuk, Andrey Savchenko, Alexander Panchenko et al.2026 · 0 citationsarXiv
SPARTA: Evaluating Reasoning Segmentation Robustness through Black-Box Adversarial Paraphrasing in Text Autoencoder Latent Space2510.24446v1 · Viktoriia Zinkovich, Anton Antonov, Andrei Spiridonov et al.2025 · 0 citationsarXiv
Emergent Misalignment via In-Context Learning: Narrow in-context examples can produce broadly misaligned LLMs2510.11288v4 · Nikita Afonin, Nikita Andriianov, Vahagn Hovhannisyan et al.2025 · 0 citationsarXiv
Bring the Apple, Not the Sofa: Impact of Irrelevant Context in Embodied AI Commands on VLA Models2510.07067v1 · Daria Pugacheva, Andrey Moskalenko, Denis Shepelev et al.2025 · 0 citationsarXiv
One Task Vector is not Enough: A Large-Scale Study for In-Context Learning2505.23911v1 · Pavel Tikhonov, Ivan Oseledets, Elena Tutubalina2025 · 0 citationsarXiv
POLAR: A Benchmark for Multilingual, Multicultural, and Multi-Event Online Polarization2505.20624v3 · Usman Naseem, Robert Geislinger, Juan Ren et al.2025 · 0 citationsarXiv
Career total: 38 works. 18 are in this corpus.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.