Latest

Every published item, in order · 8120 items
Nov 5
PapersBuildersSelected14:04
Training Data and Routing Choices Push Open Models on ARC-AGI-2

The clearest lesson is that most of the gain for the 27B model came from the serving-side inference procedure rather than fine-tuning data, which changes how practitioners should budget effort on reasoning benchmarks. Builders running small-model inference on constrained hardware can reuse the expert-pruning and KV-cache techniques directly. The study's 189-run methodology on the ARC Prize Foundation's semi-private set also offers a rare reproducible comparison for offline ARC-AGI-2 solvers.

arXiv cs.AI@Nguyen Minh Thang, Trinh Thi PhuongSource
Tomorrow 08:00, the daily lands in your inboxTelegram channel or email. Same screening as this page, no ads.
PapersBuildersSelected12:12
Systolic Linear Attention Optimises Gated Delta Attention for Google TPUs

Linear attention layers built for GPU tensor cores waste half of each TPU matrix tile, so porting code without redesign leaves large speedups on the table. Teams pre-training on TPUs can adopt SLA-style key width, gate tiling and kernel fusion to cut training time without changing perplexity. Against KDA on TPU v7x, the 11%-24% training speed gain makes this a concrete alternative for MoE runs.

arXiv cs.LG@Gagik Amirkhanyan, Nuojin ChengSource

Only the latest 2000 items are listed; for older ones see the daily archive