NightFeats Multi-Agent RAG System Wins Best Dynamic Evaluation at NeurIPS 2025 Competition
Researchers presented NightFeats, a multi-agent retrieval-augmented generation system that won the Best Dynamic Evaluation award in the text-to-text track at the MMU-RAGent competition at NeurIPS 2025. The system uses a three-phase pipeline with temporal-semantic reranking and citation-preserving composition, outperforming proprietary baselines like Claude-SonnetV2 and Nova-Pro on human evaluations. The work demonstrates that architectural transparency and evidence grounding may be better aligned with human preferences than systems optimized solely for automatic metrics.
NightFeats is a structured multi-agent retrieval-augmented generation (RAG) system designed to decompose knowledge synthesis into three coordinated phases: retrieval, curation, and composition. Each phase operates with explicit intermediate representations and handoff contracts, inspired by principles of Agentic Context Engineering (ACE). The system introduces several core architectural primitives including temporal-semantic reranking, bounded contradiction reconciliation, and citation-preserving composition. In competition results at NeurIPS 2025, NightFeats surpassed proprietary baselines including Claude-SonnetV2 and Nova-Pro on both LLM-as-a-Judge and Human Likert scale evaluations. The authors argue that their approach prioritizes architectural transparency and verifiable evidence grounding over narrow optimization for automatic similarity metrics, suggesting this alignment with human preferences represents a meaningful advance in RAG system design.
What different sources said
- arXiv cs.AICenter
NightFeats @ MMU-RAGent NeurIPS 2025: A Context-Optimized Multi-Agent RAG System for the Text-to-Text Track
Related
Genetic Drift, Not Selection, Drives Rapid Feather Color Evolution in Island Bird Radiation
A new study of an island bird radiation found that rapid evolution of feather coloration is driven primarily by genetic drift in small populations rather than sexual or ecological selection. The research integrated whole-genome data with detailed plumage measurements across complete species sampling to test whether signaling trait evolution correlates with speciation rates. The findings suggest that neutral demographic processes play a central role in generating phenotypic diversity during island radiations, challenging assumptions about the mechanisms driving rapid evolution.
New AI Model Improves Prediction of Therapeutic Peptide Function from Protein Sequences
Researchers developed a lightweight CNN classifier that predicts whether peptide sequences have therapeutic properties, trained on a database of 54,655 peptides across 48 functional categories. The model uses a novel negative sampling strategy to reduce false positive rates from over 60% in previous approaches to 2.1%. This advancement could accelerate drug discovery by enabling faster computational screening of peptide candidates before expensive experimental testing.
Study Shows Different Metabolic Stress Models Produce Distinct Effects on Human Neuronal Networks
Researchers tested three common in vitro metabolic stress models on human-derived neuronal networks and found each produced different patterns of neuronal activity and cell damage. The models tested were hypoxia alone, oxygen-glucose deprivation (OGD), and hypoxia combined with glutamate exposure. The findings suggest that choice of experimental model significantly affects results and that combining electrophysiological and structural analyses is important for accurately assessing metabolic stress in stroke research.