Alberto Aviles-Correahttps://albertoaviles.co/Recent content on Alberto Aviles-CorreaHugoen-usFri, 07 Aug 2026 00:00:00 +0000Building a Benchmark for Slay the Spirehttps://albertoaviles.co/posts/sts-bench/Fri, 07 Aug 2026 00:00:00 +0000https://albertoaviles.co/posts/sts-bench/12 LLMs attempt 36 runs of Slay the Spire through a text interface. Six of them beat the game.