一作ACL 2026Workshop· Natural Language Processing
Multi-Agent Reasoning Improves Compute Efficiency: P areto-Optimal Test-Time Scaling
Florian Valentin Wunderlich · Lars Benedikt Kaesberg · Jan Philip Wahle · Terry Ruas · Bela Gipp
摘要
Advances in inference methods have enabled language models to improve their predictions without additional training. These methods often prioritize raw performance over cost-effective compute usage. However, computational efficiency is key for real-world applications with resource constraints. We...