Post

Nature Computational Science
@natcomputsci.nature.com

📢Out now! @jixingli.bsky.social, @lamb-cityuhk.bsky.social and colleagues assess whether instruction tuning can enhance LLM's ability to capture linguistic information in the human brain. www.nature.com/articles/s43... #compneuro #ArtificialIntelligence

nature.comIncreasing alignment of large language models with language processing in the human brain - Nature Computational ScienceLarger LLMs’ self-attention more accurately predicts readers’ regressive saccades and fMRI responses in language regions, whereas instruction tuning adds no benefit.

5 reposts13 likes

Nature Computational Science
@natcomputsci.nature.com

An accompanying News & Views for this paper from Samuel A. Nastase is now available! www.nature.com/articles/s43... #compneuro #ArtificialIntelligence 🔓 rdcu.be/ePCk6

nature.comLarger language models better align with the reading brain - Nature Computational ScienceA systematic comparison of large language models suggests that larger models align better with both human behavior and brain activity during natural reading. Instruction tuning, however, does not yiel...