Arpit Sharma

Yunan Lu, Luigi Liu, Omar Yahia et al.

Evaluating LLM-based agents remains challenging because identifying meaningful failure cases often requires substantial human effort to design realistic test scenarios. Prior works primarily focus on automatically discovering agent failures induced by adversarial users, while overlooking queries with real user intents that also trigger agent failures. We introduce PQR, a framework that not only surfaces agent failures with respect to specific objectives (e.g., helpfulness, safety, etc.) but also resembles real users' intents. PQR operates through an iterative interaction between two complementary modules. The query refinement module performs rewrites to explore diverse query variations, while the prompt refinement module uses prior feedback to derive new objective-violating strategies and realism policies for refining prompts, which in turn generate failure-triggering yet realistic queries. We evaluate PQR on detecting an e-commerce QA agent's unhelpful responses. Our method uncovers 23% - 78% more unhelpful responses, and our generated queries are more diverse and realistic compared to previous methods.

Pratik Jayarao, Arpit Sharma

In this work, we retrained the distilled BERT language model for Walmart's voice shopping assistant on retail domain-specific data. We also injected universal syntactic dependencies to improve the performance of the model further. The Natural Language Understanding (NLU) components of the voice assistants available today are heavily dependent on language models for various tasks. The generic language models such as BERT and RoBERTa are useful for domain-independent assistants but have limitations when they cater to a specific domain. For example, in the shopping domain, the token 'horizon' means a brand instead of its literal meaning. Generic models are not able to capture such subtleties. So, in this work, we retrained a distilled version of the BERT language model on retail domain-specific data for Walmart's voice shopping assistant. We also included universal dependency-based features in the retraining process further to improve the performance of the model on downstream tasks. We evaluated the performance of the retrained language model on four downstream tasks, including intent-entity detection, sentiment analysis, voice title shortening and proactive intent suggestion. We observed an increase in the performance of all the downstream tasks of up to 1.31% on average.

Arpit Sharma

The Winograd Schema Challenge (WSC) is a natural language understanding task proposed as an alternative to the Turing test in 2011. In this work we attempt to solve WSC problems by reasoning with additional knowledge. By using an approach built on top of graph-subgraph isomorphism encoded using Answer Set Programming (ASP) we were able to handle 240 out of 291 WSC problems. The ASP encoding allows us to add additional constraints in an elaboration tolerant manner. In the process we present a graph based representation of WSC problems as well as relevant commonsense knowledge. This paper is under consideration for acceptance in TPLP.