CompoundingKernel
CompoundingKernel ranks #1 on SOL-ExecBench L1, L2 and Quantization boards, 4.40x faster than the reference baseline.
Self-Improving from Zero to First
Tuesday, November 17: 7 frontier papers out of 813 read.
Papers rated as moving the field
How we chooseCompoundingKernel ranks #1 on SOL-ExecBench L1, L2 and Quantization boards, 4.40x faster than the reference baseline.
Self-Improving from Zero to First
Across eight LLMs, all evaluated models exhibit safety violations on EconSafeBench, with violation rates ranging from 8.5% to 40.6%.
The 36B-A4B model nearly matches a 32B dense model trained on the same data with one-seventh the active parameters.
Optimistic verifiers are hacked by padding: a sufficiency proxy reaches 0.940 while true faithfulness falls from 0.731 to 0.667.
ScaleVLA with a 552B backbone outperforms a 2B backbone in sample efficiency, generalization, and precise manipulation.
Deceptive model organisms that anticipate auditing satisfy effectiveness, utility preservation, and robustness where prior organisms fail.
Nov 11–17: 258 cleared, 18 frontier
Papers this week that name the model
Papers this week with an author at the lab
92 papers, by field