7 papers
The More Popular, The Harder to Forget: Adaptive Popularity for LLM Unlearning
Anna Borisiuk, Andrey Savchenko, Alexander Panchenko +1
Popular facts are memorised more deeply during pretraining and resist removal longer than rare ones, yet existing LLM unlearning methods apply uniform gradient pressure regardless…
Breaking the Chain: A Causal Analysis of LLM Faithfulness to Intermediate Structures
Oleg Somov, Mikhail Chaichuk, Gleb Ershov +4
In schema-guided reasoning (SGR) pipelines, LLMs produce explicit intermediate structures -- rubrics, checklists, or verification queries -- before committing to a final decision.…
Anatomy of Unlearning: The Dual Impact of Fact Salience and Model Fine-Tuning
Anna Borisiuk, Andrey Savchenko, Alexander Panchenko +1
Machine Unlearning (MU) enables Large Language Models (LLMs) to remove unsafe or outdated information. However, existing work assumes that all facts are equally forgettable and lar…
Emergent Misalignment via In-Context Learning: Narrow in-context examples can produce broadly misaligned LLMs
Nikita Afonin, Nikita Andriianov, Vahagn Hovhannisyan +9
Recent work has shown that narrow finetuning can produce broadly misaligned LLMs, a phenomenon termed emergent misalignment (EM). While concerning, these findings were limited to f…
When Punctuation Matters: A Large-Scale Comparison of Prompt Robustness Methods for LLMs
Mikhail Seleznyov, Mikhail Chaichuk, Gleb Ershov +3
Large Language Models (LLMs) are highly sensitive to subtle, non-semantic variations in prompt phrasing and formatting. In this work, we present the first systematic evaluation of…
The benefits of query-based KGQA systems for complex and temporal questions in LLM era
Artem Alekseev, Mikhail Chaichuk, Miron Butko +3
Large language models excel in question-answering (QA) yet still struggle with multi-hop reasoning and temporal questions. Query-based knowledge graph QA (KGQA) offers a modular al…