11 citations · 11 across the 4 of their papers we have counts for
5 papers
False Floors: LLM Safety Routing Evaluations Break Under Distribution Shift
Amit Singh Bhatti, Vishal Vaddina
Safety routers send each request to one of several models and are judged against the best single model. A major routing benchmark picks that comparator on the evaluation data. In t…
PROTEUS: SLA-Aware Routing via Lagrangian RL for Multi-LLM Serving Systems
Amit Singh Bhatti, Vishal Vaddina, Dagnachew Birru
Production LLM deployments serve diverse workloads where cost and quality requirements vary by customer tier, time of day, and query criticality. Model serving systems accept laten…
Responsible AI for General-Purpose Systems: Overview, Challenges, and A Path Forward
Gourab K Patro, Himanshi Agrawal, Himanshu Gharat +4
Modern general-purpose AI systems made using large language and vision models, are capable of performing a range of tasks like writing text articles, generating and debugging codes…
ORPO-Distill: Mixed-Policy Preference Optimization for Cross-Architecture LLM Distillation
Aasheesh Singh, Vishal Vaddina, Dagnachew Birru
We introduce ORPO-Distill, a general-purpose method for cross-architecture LLM distillation that formulates the problem as a preference optimization task. Unlike standard CoT disti…
3D Conditional Generative Adversarial Networks to enable large-scale seismic image enhancement
Praneet Dutta, Bruce Power, Adam Halpert +10
We propose GAN-based image enhancement models for frequency enhancement of 2D and 3D seismic images. Seismic imagery is used to understand and characterize the Earth's subsurface f…