A Pure Indistinguishability Obfuscation Approach to Adaptively-Sound SNARGs for NP
summary
Constructs an adaptively-sound SNARG for NP in the CRS model from sub-exponential iO plus only sub-exponential one-way functions, removing the prior need for discrete-log / factoring / LWE. The construction actually builds obfuscated proof and verifier programs using puncturable PRFs, embeds a single one-way-function challenge everywhere via complexity leveraging (two-to-the-n hybrids and a longer CRS), and reduces any successful forgery to finding a preimage of an (effectively) injective one-way function, so the scheme works under the stated assumptions.
AI-written summary · about
chapters (10)
- 0:00SNARGs from iO
- 2:50Non-adaptive vs adaptive security
- 8:102014 SNARG limits
- 13:31Subexponential hardness need
- 21:33Adaptive SNARG overview
- 23:40Off-path forgeries
- 30:23One-way function challenges
- 37:56Pure iO approach
- 41:57Other SNARG works
- 43:48Q&A








