The collaborative project operates under the banner of a provably.fast workshop, with its underlying code repository hosted publicly within the starknet-innovation GitHub organization. What sets this challenge apart from traditional academic competitions is its stringent verification standard: every target statement is formally pinned in the Lean interactive theorem prover. A problem is officially recognized as solved only when accompanied by an independently reviewed, machine-checked proof, with full credit attributed directly to the individual human researcher or the AI agent responsible for the breakthrough.
To understand the significance of the challenge, one must examine the foundational technology it seeks to optimize. STARKs—which stand for Scalable Transparent Arguments of Knowledge—are a powerful class of cryptographic zero-knowledge proofs. They rely exclusively on hash functions rather than complex cryptographic assumptions like elliptic curves, making them particularly attractive for modern decentralized systems because they are transparent and post-quantum secure.
At the heart of the optimization effort lies a fundamental component of STARK verification: the Reed-Solomon proximity test. This testing mechanism determines how many queries a verifier needs to run to be confident that a given data set is close to a Reed-Solomon codeword. The number of required queries directly dictates the size of the overall proof and the time it takes to verify it. Specifically, the new challenge targets the number of what cryptographers refer to as "bad challenges" within this proximity test. By improving the theoretical bounds governing these bad challenges, researchers can directly reduce the query complexity. The current baseline for these bounds is derived from prior academic research, notably a prominent 2026 study by Dao, Kominers, and Thaler.
The project website outlines three distinct mathematical targets for participants to pursue. These objectives include replacing a currently known quadratic bound with a tighter linear one, as well as successfully proving a quadratic bound across a substantially wider range of parameters than what is currently achievable.
However, organizers emphasize that solving the mathematics on paper is only part of the battle. The project guidelines explicitly note that discovering an improved mathematical bound in isolation does not automatically shrink any proofs in practice. Real-world cryptographic savings require a comprehensive, end-to-end protocol analysis to ensure that theoretical gains translate safely into actual implementation parameters within the STARK framework.
By leveraging the Lean theorem prover, the challenge introduces a rigorous, trustless standard of mathematical validation. Lean is a proof assistant that allows mathematicians to write theorems and proofs in a formal language that can be checked by computer software for absolute correctness. This ensures that any solution submitted by human mathematicians or advanced AI agents is free from the subtle logical errors that can occasionally slip through traditional peer-review processes.
As the cryptographic and mathematical communities begin engaging with the challenge, the initiative represents a fascinating intersection of formal verification, advanced zero-knowledge cryptography, and automated reasoning. Whether the breakthrough proofs ultimately come from seasoned academic researchers or autonomous AI agents, the resulting optimizations promise to push the boundaries of how efficient and scalable zero-knowledge proofs can become.

Comments (0)