Researchers study how LLMs assign emotional value to outcomes
A new study published on arXiv examines how large language models (LLMs) process emotional responses to outcomes like gambling or stock trading. The authors tested whether models predict how outcomes will feel and whether those predictions influence decision-making. When outcomes were rated individually, models showed diminishing sensitivity to magnitude, similar to human value functions, but…
1 source primary source
Key points
- LLMs rated gains more intensely than losses, reversing human loss aversion, which researchers call silicon gain amplification
- When comparing outcomes directly, models’ emotional ratings linearized, reducing the gain-loss asymmetry in choices
- Models’ own emotional predictions did not reliably explain their decisions better than monetary utility alone
Read the original at arXiv cs.AI · by Joseph Marks (NatWest Group, London, UK), Marius C. Vollberg (University of Zurich, Zurich, Switzerland, Swiss Center for Affective Sciences, Geneva, Switzerland) primary sourceOpen source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Research
All →- SemiAnalysis releases GDPx benchmark with Claude Opus 5.5 leading at 12% · 3 src
- Gemma-3-12b fails to detect misinformation when poisoned claims dominate retrieved evidence · 1 src
- Researchers release BILAL-BENCH to test Bengali AI models’ Islamic text fabrication · 1 src
- Researchers prune 729 attention heads to cut GPT-2-XL’s memorisation by 91.5% · 1 src
- arXiv paper finds shortcuts distort RAG benchmarking accuracy · 1 src
Comments
via GitHub Discussions