We explore the use of expert iteration in the context of language modeling applied to formal
mathematics. We show that at same compute budget, expert iteration, by which we mean …

The formalization of existing mathematical proofs is a notoriously difficult process. Despite
decades of research on automation and proof assistants, writing formal proofs remains …

In theorem proving, the task of selecting useful premises from a large library to unlock the
proof of a given conjecture is crucially important. This presents a challenge for all theorem …

Mathematical reasoning poses a significant challenge for language models due to its
complex and structured nature. In this paper, we introduce DeepSeekMath 7B, which …

This paper describes the TPTP problem library and associated infrastructure, from its use of
Clause Normal Form (CNF), via the First-Order Form (FOF) and Typed First-order Form …

There is much excitement about the opportunity to harness the power of large language
models (LLMs) when building problem-solving assistants. However, the standard …

Humans prove theorems by relying on substantial high-level reasoning and problem-
specific insights. Proof assistants offer a formalism that resembles human mathematical …

The main ingredients underlying this approach are efficient automatic theorem provers that
can cope with hundreds of axioms, suitable translations of the proof assistant's logic to the …

Reasoning, a crucial ability for complex problem-solving, plays a pivotal role in various real-
world settings such as negotiation, medical diagnosis, and criminal investigation. It serves …

Sledgehammer is a component of Isabelle/HOL that employs resolution-based first-order
automatic theorem provers (ATPs) to discharge goals arising in interactive proofs. It …