EasyProver tackles Lean verification with recursive agents
The paper authors present EasyProver, a Lean proving system built with recursive agents, isolated contexts, asynchronous search, parallel tactic probing, and an elastic elaboration service. Using an off-the-shelf frontier model without additional training, they report that it proved 669 of 672 PutnamBench problems, all six IMO 2026 formalizations, 597 of 600 NTP4VC conditions, and all 189 VERINA tasks. The authors also introduce Aegis, which produces 2,045 differentially validated Rust-Lean program pairs and 11,533 candidate specifications for evaluating other provers and supplying reinforcement-learning rewards.
EasyProver combines agentic search with formal verification, while Aegis provides a new, implementation-focused evaluation resource for AI theorem provers.
Computer Science > Computation and Language
arXiv:2610.38174 (cs)
[Submitted on 28 Oct 2026]
Title:Scaling Lean Theorem Proving for Real-World Code Verification
Authors:Debargha Ganguly, Divya Raghunathan, Dejan Jovanović, Vikash Singh, Soonho Kong, Aaron Tomb, Aman Goel, Ferhat Erata, Sam Bayless
View a PDF of the paper titled Scaling Lean Theorem Proving for Real-World Code Verification, by Debargha Ganguly and 8 other authors
#EasyProver#Aegis#Lean#Aeneas#Rust#PutnamBench#NTP4VC#VERINA