Sayandeep Sen

Kishan Maharaj, Vitobha Munigala, Srikanth G. Tamilselvam et al.

Recent advancements in large language models (LLMs) have significantly enhanced their ability to understand both natural language and code, driving their use in tasks like natural language-to-code (NL2Code) and code summarisation. However, LLMs are prone to hallucination, outputs that stray from intended meanings. Detecting hallucinations in code summarisation is especially difficult due to the complex interplay between programming and natural languages. We introduce a first-of-its-kind dataset, CodeSumEval, with ~10K samples, curated specifically for hallucination detection in code summarisation. We further propose a novel Entity Tracing Framework (ETF) that a) utilises static program analysis to identify code entities from the program and b) uses LLMs to map and verify these entities and their intents within generated code summaries. Our experimental analysis demonstrates the framework's effectiveness, leading to a 73% F1 score. The proposed approach provides a method for detecting hallucinations by tracing entities from the summary to the code, allowing us to evaluate summary accuracy and localise the error within the summary.

Dushyant Behl, Palanivel Kodeswaran, Venkatraman Ramakrishna et al.

Private blockchain networks are used by enterprises to manage decentralized processes without trusted mediators and without exposing their assets publicly on an open network like Ethereum. Yet external parties that cannot join such networks may have a compelling need to be informed about certain data items on their shared ledgers along with certifications of data authenticity; e.g., a mortgage bank may need to know about the sale of a mortgaged property from a network managing property deeds. These parties are willing to compensate the networks in exchange for privately sharing information with proof of authenticity and authorization for external use. We have devised a novel and cryptographically secure protocol to effect a fair exchange between rational network members and information recipients using a public blockchain and atomic swap techniques. Using our protocol, any member of a private blockchain can atomically reveal private blockchain data with proofs in exchange for a monetary reward to an external party if and only if the external party is a valid recipient. The protocol preserves confidentiality of data for the recipient, and in addition, allows it to mount a challenge if the data turns out to be inauthentic. We also formally analyze the security and privacy of this protocol, which can be used in a wide array of practical scenarios

Vijay Arya, Sayandeep Sen, Palani Kodeswaran

There has been an unprecedented surge in the number of service providers offering a wide range of machine learning prediction APIs for tasks such as image classification, language translation, etc. thereby monetizing the underlying data and trained models. Typically, a data owner (API provider) develops a model, often over proprietary data, and leverages the infrastructure services of a cloud vendor for hosting and serving API requests. Clearly, this model assumes complete trust between the API Provider and cloud vendor. On the other hand, a malicious/buggy cloud vendor may copy the APIs and offer an identical service, under-report model usage metrics, or unfairly discriminate between different API providers by offering them a nominal share of the revenue. In this work, we present the design of a blockchain based decentralized trustless API marketplace that enables all the stakeholders in the API ecosystem to audit the behavior of the parties without having to trust a single centralized entity. In particular, our system divides an AI model into multiple pieces and deploys them among multiple cloud vendors who then collaboratively execute the APIs. Our design ensures that cloud vendors cannot collude with each other to steal the combined model, while individual cloud vendors and clients cannot repudiate their input or model executions.