Escaping Mode Collapse in LLM Generation via Geometric Regulation

arXiv:2607.20484v1 Announce Type: new Abstract: Large Language Models (LLMs) are fundamentally limited by representation collapse, a bottleneck that severely degrades long-context

arXiv:2605.19461v1 Announce Type: new Abstract: On-policy reinforcement learning methods like GRPO suffer from mode collapse: they exhibit reduced solution diversity,

arXiv:2609.38359v1 Announce Type: new Abstract: High quality synthetic data is central to post training LLMs for adaptive AI applications that represent the diverse expert

arXiv:2604.18804v1 Announce Type: cross Abstract: Latent Diffusion Models (LDMs) achieve high-fidelity synthesis but suffer from latent space brittleness, causing discontinuous

arXiv:2604.15400v1 Announce Type: cross Abstract: We present causal evidence that hallucination in autoregressive language models is an early trajectory commitment governed by

arXiv:2608.21366v1 Announce Type: new Abstract: Driven by massive amounts of web-scale data, generative AI (GenAI) has achieved remarkable progress, enabling various applications

arXiv:2606.27732v1 Announce Type: cross Abstract: Discrete diffusion language models (dLLMs) recover masked tokens in parallel, offering significant speedups over autoregressive

arXiv:2609.11146v1 Announce Type: new Abstract: AI-generated text is flowing back into the training corpora of the next generation of models. Recursive training on it drives model