Publications (7)
COBART: Controlled, Optimized, Bidirectional and Auto-Regressive Transformer for Ad Headline Generation
Yashal Shakti Kanungo, Gyanendra Das, Pooja A +1
Online ads are essential to all businesses and ad headlines are one of their core creative component. Existing methods can generate headlines automatically and also optimize their…
Fast-dLLM++: Fréchet Profile Decoding for Faster Diffusion LLM Inference
Siva Rajesh Kasa, Yasong Dai, Sumit Negi +1
Diffusion large language models promise parallel token generation, yet inference remains bottlenecked by deciding which masked tokens can be safely committed together. Fast-dLLM ad…
The Hidden Cost of Modeling P(X): Vulnerability to Membership Inference Attacks in Generative Text Classifiers
Owais Makroo, Siva Rajesh Kasa, Sumegh Roychowdhury +4
Membership Inference Attacks (MIAs) pose a critical privacy threat by enabling adversaries to determine whether a specific sample was included in a model's training dataset. Despit…
DIVERSED: Relaxed Speculative Decoding via Dynamic Ensemble Verification
Ziyi Wang, Siva Rajesh Kasa, Ankith M S +6
Speculative decoding is an effective technique for accelerating large language model inference by drafting multiple tokens in parallel. In practice, its speedup is often bottleneck…
Ad Headline Generation using Self-Critical Masked Language Model
Yashal Shakti Kanungo, Sumit Negi, Aruna Rajan
For any E-commerce website it is a nontrivial problem to build enduring advertisements that attract shoppers. It is hard to pass the creative quality bar of the website, especially…
TaTToo: Tool-Grounded Thinking PRM for Test-Time Scaling in Tabular Reasoning
Jiaru Zou, Soumya Roy, Vinay Kumar Verma +6
Process Reward Models (PRMs) have recently emerged as a powerful framework for enhancing the reasoning capabilities of large reasoning models (LRMs), particularly in the context of…
Rethinking Test Time Scaling for Flow-Matching Generative Models
Qingtao Yu, Changlin Song, Minghao Sun +6
The performance of text-to-image diffusion models may be improved at test-time by scaling computation to search for a generated image that maximizes a given reward function. While…