6 papers
ROSE: Benchmarking the Perception-to-Action Gap in Multimodal Models
Yihao Wang, Zijian He, Jie Ren +1
Multimodal large language models (MLLMs) are increasingly expected to act on visual information, yet the same scene may require different actions under different task contexts. How…
Nitsum: Serving Tiered LLM Requests with Adaptive Tensor Parallelism
Vikranth Srivatsa, Zijian He, Pu Guo +2
LLM serving is increasingly multi-tenant: the same deployment must handle latency-critical interactive requests and more relaxed background workloads under a fixed GPU budget. This…
VDCores: Resource Decoupled Programming and Execution for Asynchronous GPU
Zijian He, Adrian Sampson, Yiying Zhang +1
Modern GPUs increasingly rely on specialized and asynchronous hardware units to deliver high performance. Yet these units are often underutilized because today's GPU software stack…
ChunQiuTR: Time-Keyed Temporal Retrieval in Classical Chinese Annals
Yihao Wang, Zijian He, Jie Ren +1
Retrieval shapes how language models access and ground knowledge in retrieval-augmented generation (RAG). In historical research, the target is often not an arbitrary relevant pass…
Cognify: Supercharging Gen-AI Workflows With Hierarchical Autotuning
Zijian He, Reyna Abhyankar, Vikranth Srivatsa +1
Today's gen-AI workflows that involve multiple ML model calls, tool/API calls, data retrieval, or generic code execution are often tuned manually in an ad-hoc way that is both time…
Preble: Efficient Distributed Prompt Scheduling for LLM Serving
Vikranth Srivatsa, Zijian He, Reyna Abhyankar +2
Prompts to large language models (LLMs) have evolved beyond simple user questions. For LLMs to solve complex problems, today's practices are to include domain-specific instructions…