SakanaAI

Hypernetworks that update LLMs to remember factual information

Score basis：Clear source · Risk needs review · Universal

A Tree Search Library with Flexible API for LLM Inference-Time Scaling

Score basis：Clear source · Risk needs review · Universal

Training teachers with reinforcement learning able to make LLMs learn how to reason for test time scaling.

Score basis：Clear source · Risk needs review · Universal

Extending the Context of Pretrained LLMs by Dropping Their Positional Embedding

Score basis：Clear source · Risk needs review · Universal

Digital Red Queen: Adversarial Program Evolution in Core War with LLMs

Score basis：Clear source · Risk needs review · Universal

LLMs for Constructed Languages

Score basis：Clear source · Risk needs review · Universal

[ICLR 2026] Evaluating the performance of LLMs on Japanese challenging financial tasks.

Score basis：Clear source · Risk needs review · Universal

Cuda kernels for leveraging LLM sparsity to improve throughput and decrease the memory requirements during inference and training.

Score basis：Clear source · Risk needs review · Universal