For An Exciting Tomorrow
arXiv:2510.05942v2 Announce Type: replace-cross Abstract: We present EvalMORAAL, a transparent chain-of-thought (CoT) framework that uses two scoring methods (log-probabilities and…
arXiv:2510.07022v1 Announce Type: cross Abstract: Machine unlearning is critical for enforcing data deletion rights like the “right to be forgotten.”…
arXiv:2510.05379v2 Announce Type: replace-cross Abstract: Recent advancements in jailbreaking large language models (LLMs), such as AutoDAN-Turbo, have demonstrated the power…
arXiv:2510.05106v1 Announce Type: new Abstract: The design of safety-critical agents based on large language models (LLMs) requires more than simple…
arXiv:2510.05016v2 Announce Type: replace-cross Abstract: While task-specific demonstrations show early success in applying large language models (LLMs) to automate some…
arXiv:2510.05881v1 Announce Type: cross Abstract: We propose the Segmented Full-Song Model (SFS) for symbolic full-song generation. The model accepts a…
arXiv:2510.05862v1 Announce Type: cross Abstract: Long-context models (LCMs) have demonstrated great potential in processing long sequences, facilitating many real-world applications.…
arXiv:2510.04956v2 Announce Type: replace-cross Abstract: Computer-assisted pronunciation training (CAPT) manages to facilitate second-language (L2) learners to practice pronunciation skills by…
arXiv:2510.03285v1 Announce Type: new Abstract: Recent advances in browser-based LLM agents have shown promise for automating tasks ranging from simple…