~ similar to 2607.03561· 19 results
Xunguang Wang, Yuguang Zhou, Qingyue Wang, Zongjie Li +4 more
This paper introduces a novel framework, the Reasoning Safety Monitor, to detect and prevent logical inconsistencies and adversarial manipulations within the internal reasoning steps of large language…
The paper proves that platform-deterministic inference is a necessary and sufficient condition for trustworthy AI, establishing that AI trust fundamentally relies on consistent arithmetic.
The paper introduces a lightweight, sampling-based cryptographic protocol for verifiable AI inference that drastically reduces proving overhead from minutes to milliseconds by leveraging statistical p…
Honghao Liu, Chengjin Xu, Xuhui Jiang, Cehao Yang +4 more
The paper demonstrates that confronting Large Reasoning Models (LRMs) with conflicting objectives, such as contradictory choices or conflicting alignment values, significantly increases their vulnerab…
The paper proposes a method for bit-exact verification of AI inference outputs without sacrificing performance, demonstrating that deterministic, precise re-computation is possible even across differe…
The paper studies the resources required to batch verify Boolean functions and provides lower bounds on the witness-query tradeoff based on approximate degree.
The paper proposes a trust-boundary architecture using Lean 4 to verify the deterministic structured computations surrounding LLM pipelines, providing verifiable certificates for high-stakes deploymen…
Jek Huang, Jeffery Hsia, Jiayi Sun, Freddie Shi +2 more
This paper introduces Proof-or-Stop Lifecycle Control, a method that allows lifecycle transitions only when mechanically verifiable evidence is provided, and evaluates its implementation.
This paper establishes mathematical limits of AGI safety, proving structural unverifiability as the core barrier.
Yuxi Sun, Wenbo Shang, Wei Gao, Xin Huang +1 more
The paper introduces a diagnostic testbed, PAVE, to evaluate how LLMs arbitrate between their internal knowledge and retrieved evidence during fact-checking, revealing that this arbitration is unrelia…
The paper proposes referential security as a new paradigm for AI evaluation to ensure that safety claims and audits remain tied to specific, verifiable system instances despite continuous, unannounced…
The paper proposes a federated formal verification architecture that treats verification as a polyglot proof system, successfully validating it on complex production subsystems like a Raft consensus m…
The paper introduces containment verification, a novel method that provides safety guarantees by formally verifying the agentic framework itself, ensuring safety regardless of the underlying AI model'…
The paper introduces Oracle Poisoning, an attack that corrupts knowledge graphs used by AI agents, demonstrating that all tested models blindly trust poisoned data at high sophistication levels.
The paper analyzes the failure modes of aggressive 2-bit quantization in large reasoning models, proposing lightweight controls like FP16 planning and loop rescue to restore accuracy and achieve pract…
This paper introduces Epistemic Byzantine Fault Tolerance (EBFT), a fault-tolerance model for agentic infrastructure and post-deterministic distributed systems, addressing the Honest Quorum Problem an…