Daniel KangCan AI Speed up your Code?AI “software engineers” have been gaining a lot of attention recently. But do they actually work?Sep 3Sep 3
Daniel KangLLM Agents can Autonomously Exploit Zero-day VulnerabilitiesAgents based on large language models (LLMs) have become increasingly capable and can now solve tasks as complex as resolving real-world…Jun 51Jun 51
Daniel KangLLM Agents can Autonomously Exploit One-day VulnerabilitiesLarge language models (LLMs) have become increasingly powerful and are increasingly embodied as agents. These agents can take actions, such…Apr 161Apr 161
Daniel KangIntroducing ZKAudit — Trustless Audits of ML with zkmlMachine learning (ML) is increasingly being deployed in consequential applications — from powering algorithms that diagnose patients with…Apr 9Apr 9
Daniel KangInjecAgent: Exposing Vulnerabilities in Large Language Model AgentsAdvanced Large Language Models (LLMs) agents have the potential to serve as personal assistants in everyday life. By integrating LLMs with…Mar 11Mar 11
Daniel KangLLM Agents can Autonomously Hack WebsitesLLMs have dramatically increased in their capabilities over the past few years and can now aid in legal planning, solve Olympiad-level…Feb 13Feb 13
Daniel KangAIDB: Analyze your Unstructured Data with SQL up to 350x cheaperML is incredibly powerful, allowing analysts, social scientists, and researchers to understand the semantic contents of unstructured data…Nov 9, 20231Nov 9, 20231
Daniel KangTensorPlonk: A “GPU” for ZKML, Delivering 1,000x SpeedupsBy Suppakit Waiwitlikhit and Daniel KangSep 18, 2023Sep 18, 2023
Daniel KangFighting AI-generated Audio with Attested Microphones and ZK-SNARKs: the Attested Audio ExperimentBy Daniel Kang, Kobi Gurkan, and Anna RoseJun 13, 2023Jun 13, 2023
Daniel KangVerified Execution of GPT, Bert, CLIP, and moreIn the current era of AI-driven applications, the use of language models like GPT and BERT is pervasive. These models are the engines…May 22, 2023May 22, 2023