Tag
A commentary tweet highlighting RL with Neural Feedback (RLNF), a claimed approach to training AI models on human preferences using only brain signals, moving beyond explicit rating-based feedback.
The article discusses which jobs would survive the arrival of AGI or ASI, focusing on roles involving authority, human preference, and legal accountability.
This paper introduces RM-EVAL, a reward model trained on human preference data for reference-free meta-evaluation of grammatical error correction, and shows how it can improve GEC systems via reward-guided text generation.
A developer built a benchmark called lolbench to evaluate LLMs on understanding and creating jokes, revealing models are strong at explaining real jokes but struggle with failed ones, and incorporating human voting for preference testing.
The Artificial Analysis Image Editing Arena has been updated to expand editing tasks and use cases, providing a comprehensive leaderboard that ranks AI models on human preference for various image editing scenarios.
VA-Judger is the first reward model for joint video-audio generation that uses human preference feedback to evaluate holistic quality, including a dataset and benchmark, and demonstrates significant improvements in human preference rates when applied to models like LTX-2.
Max Planck Institute for Intelligent Systems has launched Comparity AI, a research platform for human preference-based LLM rankings that provides free access to frontier models and a personal leaderboard.
This paper presents a text-to-music generation system that leverages reward conditioning, expert iteration, and preference tuning to improve audio quality within a 120M-parameter model, submitted to the ATTM Grand Challenge at ICME 2026.
The article notes that the cost of replacing human employees with AI may be comparable to or higher than that of real employees, and humans have an inherent preference for human interaction. Thus, large-scale AI substitution faces two significant hurdles: cost and user experience.