Ramana Rao Kompella

Developing grounding techniques for LLMs poses two requirements for interactive environments, i.e., (i) the presence of rich knowledge beyond the scope of existing LLMs and (ii) the complexity of tasks that require strategic reasoning. Existing environments fail to meet both requirements due to their simplicity or reliance on commonsense knowledge already encoded in LLMs for interaction. In this article, we present PokéLLMon, a new benchmark enriched with fictional game knowledge and characterized by the intense, dynamic, and adversarial gameplay of Pokémon battles, setting new challenges for the development of grounding and reasoning techniques in interactive environments. Empirical evaluations demonstrate that existing LLMs lack game knowledge and struggle in Pokémon battles. We investigate grounding techniques that leverage feedback and game knowledge, and provide a thorough analysis of reasoning methods from a new perspective of action consistency. Additionally, we introduce higher-level reasoning challenges when playing against human players. The implementation of our benchmark is released at: https://github.com/git-disl/PokeLLMon .

This paper addresses the challenge of scaling quantum computing by employing distributed quantum algorithms across multiple processors. As quantum circuits grow in complexity, efficiently managing the entanglement resources required for communication between quantum processing units (QPUs) becomes increasingly critical. We propose novel methods to optimize entanglement consumption while accounting for network constraints. Our approach leverages graph partitioning techniques to optimize both qubit teleportation and gate teleportation, minimizing the number of Einstein-Podolsky-Rosen (EPR) pairs required to execute general quantum circuits. Additionally, we formulate an integer linear program (ILP) to further reduce entanglement requirements by mapping the logical resources of partitioned circuits to the physical constraints of the underlying quantum network. Finally, we analyze the entanglement cost for implementing the Quantum Fourier Transform (QFT) across multiple QPUs, employing an approach that exploits the circuit's structure to minimize total entanglement consumption. These contributions provide a pathway to scalable and resource-efficient distributed quantum computing.

Recently, generative models such as diffusion models (DMs) have gained prominence in various applications, and there is a growing demand for their deployment on resourceconstrained devices. Model pruning provides an effective solution by reducing the model redundancy without significantly impacting performance. However, most existing model pruning methods are designed for classification models and often lead to substantial performance degradation when applied to generative models. To address this issue, we propose Comp-Diff, a novel two stage framework of pruning and knowledge distillation tailored for diffusion models. In the pruning stage, we propose a new structured content-aware pruning (CaP) method within CompDiff to identify and preserve informative units (filters/channels) that actually contribute to the generative capability of the model. Specifically, we introduce input perturbations to the pre-trained model and measure each unit's importance score using gradients induced by these perturbations. Units with higher importance scores are considered more informative and are retained to maintain the model's generative power. In the fine-tuning stage of Comp-Diff, we propose the distribution-aware knowledge distillation (DaKD) method, which effectively transfers finegrained knowledge from the original model to the pruned one on both attention and noise distribution levels. In addition, DaKD includes an adversarial loss to improve the quality and diversity of generated outputs. To verify and evaluate our method, we apply the proposed Comp-Diff on three representative tasks: unconditional image generation, conditional image generation, and text-to-image generation. Extensive experiments on both multi-step and one-step diffusion models demonstrate that the proposed framework consistently yields compact models and outperforms existing pruning techniques by a large margin.