2 papers
cs.LG2026
Characterizing Paraphrase-Induced Failures in Lean 4 Autoformalization
William Feng, Ethan Lou, Aryan Sharma
Lean 4 autoformalization has become increasingly popular in recent years, with frontier language models and open-weight autoformalizers now producing valid formalizations of mathem…
cs.CL2026
H-Probes: Extracting Hierarchical Structures From Latent Representations of Language Models
Cutter Dawes, Aryan Sharma, Angelos Ioannis Lagos +1
Representing and navigating hierarchy is a fundamental primitive of reasoning. Large language models have demonstrated proficiency in a wide variety of tasks requiring hierarchical…