arXiv:2604.22786v1 Announce Type: new Abstract: We present AutoCompress, a transformer compression method motivated by an empirical finding: in small transformers, Layer 0 carries disproportionately high task-critical information, with an NTK-based importance score of 3.6 compared to a maximum of 0.054…
arXiv:2604.24545v1 Announce Type: cross Abstract: In many areas of medicine, security, and life sciences, we want to allocate limited resources to different sources in order to detect extreme values. In this paper, we study an efficient way to allocate these…
A new study reveals cellular pathways that appear to underlie some differences in physical fitness.
New research suggests constructing a simple building from interlocking subunits should be mechanically feasible and have a much smaller carbon footprint.
arXiv:2511.21715v2 Announce Type: replace-cross Abstract: This paper argues that dataset structure is important in image recognition tasks (among other tasks). Specifically, we focus on the nature and genesis of correlational structure in the actual datasets upon which DNNs are trained.…
arXiv:2604.16514v4 Announce Type: replace-cross Abstract: Autoregressive vision-language models (VLMs) deliver strong multimodal capability, but their token-by-token decoding imposes a fundamental inference bottleneck. Diffusion VLMs offer a more parallel decoding paradigm, yet directly converting a pretrained autoregressive VLM into a large-block…
arXiv:2603.11504v2 Announce Type: replace Abstract: Recent reasoning models such as OpenAI-o1 and DeepSeek-R1 have shown strong performance on complex tasks including mathematical reasoning and code generation. However, this performance gain comes with substantially longer output sequences, leading to significantly increased…
A new study reveals cellular pathways that appear to underlie some differences in physical fitness.
New research suggests constructing a simple building from interlocking subunits should be mechanically feasible and have a much smaller carbon footprint.
arXiv:2511.21715v2 Announce Type: replace-cross Abstract: This paper argues that dataset structure is important in image recognition tasks (among other tasks). Specifically, we focus on the nature and genesis of correlational structure in the actual datasets upon which DNNs are trained.…
