2 citations · 2 across the 16 of their papers we have counts for
16 papers
Break It Down, Pass It On: Cross-Task Skill Transfer in LLM Agents
Yiyang Feng, Biddut Sarker Bijoy, Niranjan Balasubramanian +1
Large language model (LLM) agents can induce skills from completed tasks and reuse them later to grow more capable with experience. In practice, induced skills may transfer unrelia…
Certificates without Electrons? Theory and Evidence on Impacts from AI-Driven Power Demand
Dana Golden, Aruna Balasubramanian, Niranjan Balasubramanian
Data centers now account for 4.4% of United States electricity demand, yet the grid-level effectiveness of the renewable energy certificates (RECs) and power purchase agreements (P…
AppWorld-UL: Benchmarking Diverse Agent-User Interactions for Tool-Use
Junzhi Chen, Harsh Trivedi, Jane Pan +4
Tool-use agents that address day-to-day digital tasks such as ordering groceries must not only operate applications, but also interact with the user, e.g., to ask clarification que…
Syntax Is Easy, Semantics Is Hard: Evaluating LLMs for LTL Translation
Priscilla Kyei Danso, Mohammad Saqib Hasan, Niranjan Balasubramanian +1
Propositional Linear Temporal Logic (LTL) is a popular formalism for specifying desirable requirements and security and privacy policies for software, networks, and systems. Yet ex…
Teaching and Evaluating LLMs to Reason About Polymer Design Related Tasks
Dikshya Mohanty, Mohammad Saqib Hasan, Syed Mostofa Monsur +3
Research in AI4Science has shown promise in many science applications, including polymer design. However, current LLMs are ineffective in this problem space because: (i) most model…
Fine-Grained Energy Prediction For Parallellized LLM Inference With PIE-P
Anurag Dutt, Young Won Choi, Avirup Sil +3
With the widespread adoption of Large Language Models (LLMs), energy costs of running LLMs is quickly becoming a critical concern. However, precisely measuring the energy consumpti…
Handling Open-Vocabulary Constructs in Formalizing Specifications: Retrieval-Augmented Parsing with Expert Knowledge
Mohammad Saqib Hasan, Sayontan Ghosh, Dhruv Verma +4
We study the problem of Open-Vocabulary Constructs(OVCs) -- ones not known beforehand -- in the context of converting natural language (NL) specifications into formal languages (e.…
ProST: Progressive Sub-task Training for Pareto-Optimal Multi-agent Systems Using Small Language Models
Biddut Sarker Bijoy, Mohammad Saqib Hasan, Pegah Alipoormolabashi +3
Multi-agent systems with smaller language models (SLMs) present a viable alternative to single agent systems powered by large language models (LLMs) for addressing complex problems…
MuSciClaims: Multimodal Scientific Claim Verification
Yash Kumar Lal, Manikanta Bandham, Mohammad Saqib Hasan +3
Assessing scientific claims requires identifying, extracting, and reasoning with multimodal data expressed in information-rich figures in scientific literature. Despite the large b…
Quantifying Misattribution Unfairness in Authorship Attribution
Pegah Alipoormolabashi, Ajay Patel, Niranjan Balasubramanian
Authorship misattribution can have profound consequences in real life. In forensic settings simply being considered as one of the potential authors of an evidential piece of text o…
Teaching an Old LLM Secure Coding: Localized Preference Optimization on Distilled Preferences
Mohammad Saqib Hasan, Saikat Chakraborty, Santu Karmaker +1
LLM generated code often contains security issues. We address two key challenges in improving secure code generation. First, obtaining high quality training data covering a broad s…
: A Dataset for ynamic nformation nd ental modeling f umeric iscussions
Sayontan Ghosh, Mahnaz Koupaee, Yash Kumar Lal +4
Understanding multiparty conversations demands robust Theory of Mind (ToM) capabilities, including the ability to track dynamic information, manage knowledge asymmetries, and disti…
AppWorld: A Controllable World of Apps and People for Benchmarking Interactive Coding Agents
Harsh Trivedi, Tushar Khot, Mareike Hartmann +6
Autonomous agents that address day-to-day digital tasks (e.g., ordering groceries for a household), must not only operate multiple apps (e.g., notes, messaging, shopping app) via A…
On Initializing Transformers with Pre-trained Embeddings
Ha Young Kim, Niranjan Balasubramanian, Byungkon Kang
It has become common practice now to use random initialization schemes, rather than the pre-trained embeddings, when training transformer based models from scratch. Indeed, we find…
Look Hear: Gaze Prediction for Speech-directed Human Attention
Sounak Mondal, Seoyoung Ahn, Zhibo Yang +4
For computer systems to effectively interact with humans using spoken language, they need to understand how the words being generated affect the users' moment-by-moment attention.…
CaT-BENCH: Benchmarking Language Model Understanding of Causal and Temporal Dependencies in Plans
Yash Kumar Lal, Vanya Cohen, Nathanael Chambers +2
Understanding the abilities of LLMs to reason about natural language plans, such as instructional text and recipes, is critical to reliably using them in decision-making systems. A…