Multi-author style detection with stylistic embeddings
We investigate the use of Large Language Models (LLMs) in multi-author style change detection.
ProofPaper ↗
Key points
- Using off-the-shelf LLMs or supervised fine-tuning solely within the target domain fails to yield effective features; for this reason, we propose cross-domain transfer learning with LLM-based embeddings fine-tuned for authorship verification.
- We fine-tune Mistral-7B embedding model and compare it with DeBERTa, used in the state-of-the-art (SOTA) systems in PAN-25 multi-author writing style analysis challenge.
- Results show that DeBERTa and Mistral excel in different user-generated text scenarios.
- We exploit this diversity by combining both embeddings, leading to more effective style change detection.
Sources (1)
- [1]Multi-author style detection with stylistic embeddingsarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 16, 08:01 PM
We investigate the use of Large Language Models (LLMs) in multi-author style change detection.
Using off-the-shelf LLMs or supervised fine-tuning solely within the target domain fails to yield effective features; for this reason, we propose cross-domain transfer learning with LLM-based embeddings fine-tuned for authorship verification.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 16, 2026Perceive, Solve, Reason: Progressive Capability Formation in Time Series-LLMs
- Oct 16, 2026NL2SQL with Iterative Self-Refinement
- Oct 15, 2026CAT-alyzing Honesty: Continuous Adversarial Training for Transferable LLM Honesty
- Oct 15, 2026Name-Bias in Large Language Models
- Oct 15, 2026Incentivizing Tool-Use Reinforcement Learning for Post-Arrival Emergency Care Prediction
- Oct 15, 2026VALCON: Benchmarking and Enhancing Value Consistency in Large Language Models via Dual-Role Variation