Model releases, architecture work and prompting research on large language models — from frontier-lab announcements to the arXiv papers behind them.
arXiv:2608. 08148v1 Announce Type: cross Abstract: Attention mechanisms have been widely utilized in modern deep learning, and many existing multi-omics models inherit their conventional use to allow unrestricted bidirectional interactions.
arXiv:2608. 08195v1 Announce Type: cross Abstract: Large language models (LLMs) are high-value assets that can be derived through redeployment, fine-tuning, quantization, or further alignment.
arXiv:2608. 08676v1 Announce Type: cross Abstract: Semantic vision encoders have become a central visual interface for multimodal understanding and semantic conditioning in image generation.
arXiv:2608. 08684v1 Announce Type: cross Abstract: Long-context LLM inference is bottlenecked by KV cache memory, yet distributing a limited cache budget across layers remains challenging.
arXiv:2608. 08814v1 Announce Type: cross Abstract: We present 360CityArena, a benchmark for evaluating the urban exploration capabilities of embodied agents within a photorealistic environment constructed from 360-degree videos.
arXiv:2608. 09605v1 Announce Type: cross Abstract: Large Language Models (LLMs) have emerged as powerful tools for improving recommendation systems.
arXiv:2608. 09732v1 Announce Type: cross Abstract: Agent skills are emerging as an important attack surface in LLM-based agent systems.
arXiv:2608. 09818v1 Announce Type: cross Abstract: Reliable medical image understanding requires models to connect clinical language and visual reasoning with pixel-level grounding.
arXiv:2510. 27544v3 Announce Type: replace Abstract: Current training paradigms, optimized for long-horizon reasoning trace execution, have made Large Language Models (LLMs) excel at pattern matching and forward simulation of reasoning, but underperform at counterfactual causal understanding and reasoning.
arXiv:2607. 28432v2 Announce Type: replace Abstract: Intelligence is commonly understood as the ability to acquire and apply knowledge, adapt to unfamiliar situations and solve new problems.
arXiv:2408. 04619v2 Announce Type: replace-cross Abstract: The Transformer architecture underpins modern large language models powering state-of-the-art text generation and AI applications.
arXiv:2506. 19125v2 Announce Type: replace-cross Abstract: Although transformers are most commonly optimized as statistical sequence models, it is unclear to what extent they can implement and learn exact algorithmic computations.
arXiv:2605. 16902v2 Announce Type: replace Abstract: Scientific artifacts such as models and datasets are foundations for research.
arXiv:2607. 10202v2 Announce Type: replace Abstract: Forced-choice probes with counterbalanced orientations are a standard tool for measuring language-model "value dispositions," and a concentration/extremity index over repeated draws is read as how sharply a model commits.
arXiv:2604. 23333v2 Announce Type: replace Abstract: Scaling test-time computation with reinforcement learning (RL) has emerged as a reliable path to improve large language models (LLM) reasoning ability.
arXiv:2605. 17062v3 Announce Type: replace-cross Abstract: Spracklen et al.
arXiv:2605. 22223v2 Announce Type: replace Abstract: We study how we can leverage only a handful of characteristics of a transformer's architecture to closely predict the number of different sequences it can output, both qualitatively and quantitatively.
arXiv:2603. 00040v3 Announce Type: replace-cross Abstract: Achieving reliable 4-bit attention is a prerequisite for end-to-end FP4 computation on emerging FP4-capable GPUs, yet attention remains the main obstacle due to FP4's tiny dynamic range and attention's heavy-tailed activations.
arXiv:2604. 20244v2 Announce Type: replace-cross Abstract: Knowledge distillation (KD) is a powerful paradigm for compressing large language models (LLMs), whose effectiveness depends on intertwined choices of divergence direction, optimization strategy, and data regime.
arXiv:2605. 12991v3 Announce Type: replace-cross Abstract: LLM-based multi-agent pipelines flip from correct to incorrect answers under simulated peer disagreement at rates we term yield, a vulnerability widely attributed to RLHF-induced sycophancy.