Recent reasoning paper from Apple is an interesting study, & getting much attention (e.g., article below), but.... (1/2) www.theguardian.com/technology/2... - ThreadSky

About ThreadSky

melaniemitchell.bsky.social • 7 days ago

Recent reasoning paper from Apple is an interesting study, & getting much attention (e.g., article below), but.... (1/2)

https://www.theguardian.com/technology/2025/jun/09/apple-artificial-intelligence-ai-study-collapse

Comments

melaniemitchell.bsky.social•7 days ago

...it basically confirmed what is already well-established: LLMs (& LRMs & "LLM agents") have trouble w/ problems that require many steps of reasoning/planning.

See, e.g., lots of recent papers by Subbarao Kambhampati's group at ASU. (2/2)

danieljbutler.bsky.social•7 days ago

Do we know if the number of steps they can perform is related to how many steps they saw in their training data? Can RL fine-tuning increase the number of steps?