`lf-lean`: The frontier of verified software engineering

Thu, 05 Feb 2026 00:00:00 +0000

Introduction

As AIs automate increasingly complex software tasks, one question gets harder to answer: how do we know the code they produce is correct? The standard approach of reviewing AI-generated code and its tests doesn’t scale. Human review effort grows proportionally with code volume, while AI code generation capacity grows exponentially.1 1 Frontier software engineering typically results in multiplicative bug growth; many bugs are subtle and interaction-driven, and their count grows super-linearly with code length and complexity. See Appendix C. If this trend continues, we’re headed toward a world where code deployment is bottlenecked on human review capacity, or just deployed without review.

Catching bugs with fractional proofs

Tue, 07 Oct 2025 00:00:00 +0000

Fractional proof decomposition fuses partial evaluation and property-based testing to scale testing compute logarithmically with bug rarity, instead of linearly.

Introduction

Most large projects have a limited compute budget for testing, lest it slow down the CI pipeline. This means missing rare edge cases until customers find them in production. I love formal verification! And I want to bring the power of reasoning about program structure to testing. In this blog post, I will briefly introduce fractional proof decomposition, our new technique that makes testing compute scale logarithmically with bug rarity instead of linearly.

As a demo, I’ll walk through how this approach would have caught Anthropic’s recent approximate top-K bug. Note that we generate these unit tests without relying on the bug reproducer code, using just the high-level specification of top-K sampling.

Blog on Theorem

`lf-lean`: The frontier of verified software engineering

Introduction

Catching bugs with fractional proofs

Introduction