5 papers
Auditing Reward Hackability in Code RL Training Environments
Shreshth Rajan
We measure the rate at which code RL environments accept incorrect solutions as correct. On a 49-task sample of SWE-bench Verified, 28.5% of tasks have test suites weak enough that…
MultiVer: Zero-Shot Multi-Agent Vulnerability Detection
Shreshth Rajan
We present MultiVer, a zero-shot multi-agent system for vulnerability detection that achieves state-of-the-art recall without fine-tuning. A four-agent ensemble (security, correctn…
Uncertainty-Gated Region-Level Retrieval for Robust Semantic Segmentation
Shreshth Rajan, Raymond Liu
Semantic segmentation of outdoor street scenes plays a key role in applications such as autonomous driving, mobile robotics, and assistive technology for visually-impaired pedestri…
Arithmetic-Intensity-Aware Quantization
Taig Singh, Shreshth Rajan, Nikhil Jain
As modern neural networks become increasingly memory-bound, inference throughput is limited by DRAM bandwidth rather than compute. We present Arithmetic-Intensity-Aware Quantizatio…
Multi-Agent Code Verification via Information Theory
Shreshth Rajan
LLMs generate buggy code: 29.6% of SWE-bench solved patches fail, 62% of BaxBench solutions have vulnerabilities, and existing tools only catch 65% of bugs with 35% false positives…