arXiv:2604.10079v3 Announce Type: replace
Abstract: Supervised Fine-Tuning (SFT) is the standard approach for adapting large language models (LLMs) to downstream tasks. However, we observe a persistent failure mode: even after convergence, models ofte…
arXiv:2505.11336v4 Announce Type: replace
Abstract: Despite the growing adoption of large language models (LLMs) in academic workflows, their capabilities remain limited in supporting high-quality scientific writing. Most existing systems are designed…
arXiv:2604.03873v3 Announce Type: replace-cross
Abstract: Black-box knowledge distillation for large language models presents a strict trade-off. Simple off-policy methods (e.g., sequence-level knowledge distillation) struggle to correct the student’s…
arXiv:2509.23649v2 Announce Type: replace-cross
Abstract: Generative recommendation, which directly generates item identifiers, has emerged as a promising paradigm for recommendation systems. However, its potential is fundamentally constrained by the …
arXiv:2604.10072v3 Announce Type: replace
Abstract: Recent advancements in the Generative Reward Model (GRM) have demonstrated its potential to enhance the reasoning abilities of LLMs through Chain-of-Thought (CoT) prompting. Despite these gains, exis…
arXiv:2512.05591v2 Announce Type: replace-cross
Abstract: Large language model post-training relies on reinforcement learning to improve model capability and alignment quality. However, the off-policy training paradigm introduces distribution shift, w…
arXiv:2604.21308v1 Announce Type: cross
Abstract: Enterprise LLM agents can dramatically improve workplace productivity, but their core capability, retrieving and using internal context to act on a user’s behalf, also creates new risks for sensitive i…
arXiv:2604.21045v1 Announce Type: new
Abstract: Simultaneous speech translation (SST) generates translations while receiving partial speech input. Recent advances show that large language models (LLMs) can substantially improve SST quality, but at the…
arXiv:2604.21327v1 Announce Type: cross
Abstract: Test-time reinforcement learning (TTRL) always adapts models at inference time via pseudo-labeling, leaving it vulnerable to spurious optimization signals from label noise. Through an empirical study, …
arXiv:2604.15827v2 Announce Type: replace-cross
Abstract: Conventional information retrieval is concerned with identifying the relevance of texts for a given query. Yet, the conventional definition of relevance is dominated by aspects of similarity in…