5 papers
Many Right Answers, One Training Example: Removing Responder Noise from GRPO for LLM Recommenders
Xin Li, Hao Jiang, Jinghao Guo +2
A recommendation request on Reddit draws several replies, each naming different titles. When an LLM recommender is trained on such threads with reinforcement learning (RL), which r…
hacktrace: behavior-supervised detection of reward hacking during code generation
Hao Jiang, Xin Li, Annan Wang +2
A coding agent can earn a passing grade by fixing its code, or by deleting the test that exposes the bug. Detecting such reward hacking requires recognizing attempted shortcuts, in…
PRISMR: Overcoming Parse Collapse in Multimodal Listwise Ranking via Parameterized Representation Internalization
Hao Jiang, Xin Li, Annan Wang +4
Generative listwise ranking with Large Multimodal Models (LMMs) aims to capture global list context in a single forward pass, but its effectiveness degrades in long-context multimo…
RLPO: Residual Listwise Preference Optimization for Long-Context Review Ranking
Hao Jiang, Zhi Yang, Annan Wang +2
Review ranking is pivotal in e-commerce for prioritizing diagnostic and authentic feedback from the deluge of user-generated content. While large language models have improved sema…
MRSE: An Efficient Multi-modality Retrieval System for Large Scale E-commerce
Hao Jiang, Haoxiang Zhang, Qingshan Hou +4
Providing high-quality item recall for text queries is crucial in large-scale e-commerce search systems. Current Embedding-based Retrieval Systems (ERS) embed queries and items int…