Execution envelopes define a shared admission contract for backend AI execution requests to attach unified logging, governance, and policy hooks at entry.
Title resolution pending
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
years
2026 4representative citing papers
PASS replaces private-key control in blockchain wallets with provenance tracking, using an Inbox-Outbox mechanism for verifiable lineage on external actions while preserving privacy for internal transfers, with Lean 4 proofs and an enclave-based prototype.
An LLM-based generator-critic loop autoformalizes natural language policies into Cedar policies that cover substantially more of the source specification than hand-coded symbolic enforcement on MedAgentBench.
RLVR training raises verified Dafny pass rates from 9.7% to 31.1% on a filtered benchmark while a Lean proof scaffold lifts success from 46.2% to 69.2% on a pilot set and solves 7 of 42 prior unsolved tasks.
citing papers explorer
-
Execution Envelopes: A Shared Admission Contract for Backend AI Execution Requests
Execution envelopes define a shared admission contract for backend AI execution requests to attach unified logging, governance, and policy hooks at entry.
-
PASS: A Provenanced Access Subaccount System for Blockchain Wallets
PASS replaces private-key control in blockchain wallets with provenance tracking, using an Inbox-Outbox mechanism for verifiable lineage on external actions while preserving privacy for internal transfers, with Lean 4 proofs and an enclave-based prototype.
-
Autoformalization of Agent Instructions into Policy-as-Code
An LLM-based generator-critic loop autoformalizes natural language policies into Cedar policies that cover substantially more of the source specification than hand-coded symbolic enforcement on MedAgentBench.
-
Automating Formal Verification with Reinforcement Learning and Recursive Inference
RLVR training raises verified Dafny pass rates from 9.7% to 31.1% on a filtered benchmark while a Lean proof scaffold lifts success from 46.2% to 69.2% on a pilot set and solves 7 of 42 prior unsolved tasks.