AI RESEARCH
The Quantization Trap: Breaking Linear Scaling Laws in Multi-Hop Reasoning
arXiv CS.AI
•
ArXi:2602.13595v2 Announce Type: replace Neural scaling laws provide a predictable recipe for AI advancement: reducing numerical precision should linearly improve computational efficiency and energy profile ($E \propto \mathrm{bits}$). In this paper, we nstrate that this scaling law breaks in the context of multi-hop reasoning. We reveal a 'quantization trap' where reducing precision from 16-bit to 8/4-bit paradoxically increases net energy consumption while degrading reasoning accuracy.