arXiv subjects

Angelina Parfenova

Publications and source records attributed to Angelina Parfenova.

6 recordsLinked to original sources

arXiv2026

Do Chess Explanations Reflect Model Decisions? Behavioral and Token-Level Tests of LLM Reasoning Faithfulness

Large language models can produce fluent explanations for chess moves, but plausible language does not necessarily reflect the reasoning behind a decision. We study this question in chess, where the board state is fully observable, legal actions can be enumerated, and move quality can be evaluated independently. Across 200 Lichess endgame puzzles, we test explanations using move recoverability, decoder-side controls, and token-level scoring of legal candidate moves. Unmasked explanations make generated moves easy to recover, but this advantage drops sharply after explicit move hints are removed. Under strict masking, explanations provide only small and decoder-dependent gains over the board state alone. Token-level scoring shows that explanations can nevertheless alter move preferences: random but plausible explanations from other puzzles reduce the probability of the correct move, indicating that irrelevant reasoning text is not simply ignored. We also find that recognizable endgame motifs can make generated moves easier to recover without reliably improving move correctness. Together, these results show that linguistic plausibility, consistency with a generated action, and solution correctness are distinct properties. Fluent chess explanations can influence action preferences and support a coherent move narrative while providing only limited evidence of faithful reasoning.

cs.CL↗

arXiv2026

From Quotes to Concepts: Axial Coding of Political Debates with Ensemble LMs

Axial coding is a commonly used qualitative analysis method that enhances document understanding by organizing sentence-level open codes into broader categories. In this paper, we operationalize axial coding with large language models (LLMs). Extending an ensemble-based open coding approach with an LLM moderator, we add an axial coding step that groups open codes into higher-order categories, transforming raw debate transcripts into concise, hierarchical representations. We compare two strategies: (i) clustering embeddings of code-utterance pairs using density-based and partitioning algorithms followed by LLM labeling, and (ii) direct LLM-based grouping of codes and utterances into categories. We apply our method to Dutch parliamentary debates, converting lengthy transcripts into compact, hierarchically structured codes and categories. We evaluate our method using extrinsic metrics aligned with human-assigned topic labels (ROUGE-L, cosine, BERTScore), and intrinsic metrics describing code groups (coverage, brevity, coherence, novelty, JSD divergence). Our results reveal a trade-off: density-based clustering achieves high coverage and strong cluster alignment, while direct LLM grouping results in higher fine-grained alignment, but lower coverage 20%. Overall, clustering maximizes coverage and structural separation, whereas LLM grouping produces more concise, interpretable, and semantically aligned categories. To support future research, we publicly release the full dataset of utterances and codes, enabling reproducibility and comparative studies.

cs.CL↗

arXiv2025

Emergent Convergence in Multi-Agent LLM Annotation

Large language models (LLMs) are increasingly deployed in collaborative settings, yet little is known about how they coordinate when treated as black-box agents. We simulate 7500 multi-agent, multi-round discussions in an inductive coding task, generating over 125000 utterances that capture both final annotations and their interactional histories. We introduce process-level metrics: code stability, semantic self-consistency, and lexical confidence alongside sentiment and convergence measures, to track coordination dynamics. To probe deeper alignment signals, we analyze the evolving geometry of output embeddings, showing that intrinsic dimensionality declines over rounds, suggesting semantic compression. The results reveal that LLM groups converge lexically and semantically, develop asymmetric influence patterns, and exhibit negotiation-like behaviors despite the absence of explicit role prompting. This work demonstrates how black-box interaction analysis can surface emergent coordination strategies, offering a scalable complement to internal probe-based interpretability methods.

cs.CL↗

arXiv2025

Text Annotation via Inductive Coding: Comparing Human Experts to LLMs in Qualitative Data Analysis

This paper investigates the automation of qualitative data analysis, focusing on inductive coding using large language models (LLMs). Unlike traditional approaches that rely on deductive methods with predefined labels, this research investigates the inductive process where labels emerge from the data. The study evaluates the performance of six open-source LLMs compared to human experts. As part of the evaluation, experts rated the perceived difficulty of the quotes they coded. The results reveal a peculiar dichotomy: human coders consistently perform well when labeling complex sentences but struggle with simpler ones, while LLMs exhibit the opposite trend. Additionally, the study explores systematic deviations in both human and LLM generated labels by comparing them to the golden standard from the test set. While human annotations may sometimes differ from the golden standard, they are often rated more favorably by other humans. In contrast, some LLMs demonstrate closer alignment with the true labels but receive lower evaluations from experts.

cs.CL↗

arXiv2024

Risk prediction of pathological gambling on social media

This paper addresses the problem of risk prediction on social media data, specifically focusing on the classification of Reddit users as having a pathological gambling disorder. To tackle this problem, this paper focuses on incorporating temporal and emotional features into the model. The preprocessing phase involves dealing with the time irregularity of posts by padding sequences. Two baseline architectures are used for preliminary evaluation: BERT classifier on concatenated posts per user and GRU with LSTM on sequential data. Experimental results demonstrate that the sequential models outperform the concatenation-based model. The results of the experiments conclude that the incorporation of a time decay layer (TD) and passing the emotion classification layer (EmoBERTa) through LSTM improves the performance significantly. Experiments concluded that the addition of a self-attention layer didn't significantly improve the performance of the model, however provided easily interpretable attention scores. The developed architecture with the inclusion of EmoBERTa and TD layers achieved a high F1 score, beating existing benchmarks on pathological gambling dataset. Future work may involve the early prediction of risk factors associated with pathological gambling disorder and testing models on other datasets. Overall, this research highlights the significance of the sequential processing of posts including temporal and emotional features to boost the predictive power, as well as adding an attention layer for interpretability.

cs.CL↗

arXiv2024

Automating the Information Extraction from Semi-Structured Interview Transcripts

This paper explores the development and application of an automated system designed to extract information from semi-structured interview transcripts. Given the labor-intensive nature of traditional qualitative analysis methods, such as coding, there exists a significant demand for tools that can facilitate the analysis process. Our research investigates various topic modeling techniques and concludes that the best model for analyzing interview texts is a combination of BERT embeddings and HDBSCAN clustering. We present a user-friendly software prototype that enables researchers, including those without programming skills, to efficiently process and visualize the thematic structure of interview data. This tool not only facilitates the initial stages of qualitative analysis but also offers insights into the interconnectedness of topics revealed, thereby enhancing the depth of qualitative analysis.

cs.CL↗