#selection

Knowledge of graders can mediate selection for grader-targeting reasoning through RL

TL;DR: We insert knowledge of specific grading processes via SDF. We then show that models RL’d on environments with these graders exhibit grader-targeting reasoning, and generalise to optimising held-out environments with graders that were described in the SDF mix but not trained on during RL. Figure 1: Outline of our experiments. (Left) We use SDF to instil knowledge of grading on different subsets of environments. (Middle) We then perform RL on one half of the set of environments, which are In-Distribution (ID) for RL. We observe that models with ID grader knowledge exhibit faster learning

LessWrong · EN

Linux Fixes Faulty Mitigation Selection For BHI Vulnerability

This week's batch of x86 fixes have been sent out ahead of the Linux 7.4-rc2 kernel releasing later today...

Phoronix · EN

Compilation 37

A new selection of geometric and psychedelic illustrations by YoAz. Bold color relationships, intricate linework and carefully balanced compositions turn familiar shapes into vivid graphic worlds. This collection reflects an ongoing exploration of symmetry, rhythm and visual storytelling through contemporary illustration.

Behance Featured · EN