Improving Model Safety Behavior with Rule-Based Rewards
We've developed and applied a new method leveraging Rule-Based Rewards (RBRs) that aligns models to behave safely without extensive human data collection.
Insights
Engineering perspectives on architecture, product development, AI, and the everyday decisions behind useful software.
From the engineering desk
Browse practical articles or follow the latest technology updates.
Subscribe via RSS →Curated links from external sources — not 360Softy original articles.
We've developed and applied a new method leveraging Rule-Based Rewards (RBRs) that aligns models to behave safely without extensive human data collection.
Turbopack is a new JavaScript/TypeScript bundler we’ve been cooking at Vercel. Building on 10+ years of learnings from webpack, we want to build a bundler that can be used with many frameworks. We’re moving the Turbopack codebase into the Next.js repository—and wanted to share an update on our progress with Turbopack so far, as well as where we’re headed. Read more
Web rendering has evolved from simple server-rendered HTML pages to highly interactive and dynamic applications, and there are more ways than ever to present your app to users. Static Site Generation (SSG), Server-Side Rendering (SSR), Client-Side Rendering (CSR), Incremental Static Regeneration (ISR), and experimental Partial Prerendering (PPR) have all been developed to optimize performance, SEO, and user experience in various situations. Read more
How do you run non-blocking schema changes in MySQL? This is an eternal question. With a plethora of 3rd party solutions and with recent advancements in MySQL, it's difficult to track which solution is preferable for a given schema migration. In this post, we provide a high level overview of the state of MySQL online schema migrations in 2024. We limit the discussion to ALTER TABLE statements, as other DDL statements are typically fast (DROP TABLE is somewhat of an exception, but out of scope of
Which ORM is the fastest? We have created open-source performance benchmarks to compare query latencies for Prisma ORM, TypeORM and Drizzle ORM.
Introduction I recently encountered an intriguing bug.A user reported that their query was causing VTGate to fetch a large amount of data, sometimes resulting in an Out Of Memory (OOM) error.For a deeper understanding of grouping and aggregations on Vitess, I recommend reading this prior blog post. The Query The problematic query was:select sum(user.type) from user join user_extra on user.team_id = user_extra.id group by user_extra.id order by user_extra.id; The planner was unable to delega
Let’s start with a conversation
An idea, a challenge, or a system that needs to work better. We’ll help you understand the next step.
Prefer email? [email protected]