Papers by Yuanbo Zhang

9 papers
Synthesizing and Adapting Error Correction Data for Mobile Large Language Model Applications (2025.acl-industry)

Copied to clipboard

Challenge: Recent advances in large language models (LLMs) have achieved impressive performance on many language tasks.
Approach: They synthesize a high-quality dataset of error correction pairs to evaluate and improve LLMs for mobile applications by reweighting the sample.
Outcome: The proposed model improves on offline evaluation and live A/B testing, given the LLM performance on offline data and scores from a small privacy-preserving on-device language model.
Federated Learning of Gboard Language Models with Differential Privacy (2023.acl-industry)

Copied to clipboard

Challenge: Using federated learning and differential privacy, we train and deploy language models with federation and DP in Google Keyboard.
Approach: They train and deploy language models with federated learning and differential privacy in Google Keyboard .
Outcome: The proposed algorithm achieves meaningfully formal DP guarantees without uniform sampling of clients.
QiMeng-PRepair: Precise Code Repair via Edit-Aware Reward Optimization (2026.acl-long)

Copied to clipboard

Challenge: Existing approaches to program repair are based on correctness alone.
Approach: They propose a framework that mitigates over-editing and improves repair accuracy by generating buggy programs and re-edits.
Outcome: The proposed framework improves repair precision by 31.4% under fix1@1, a metric that considers repair correctness and extent, and significantly increases decoding throughput when combined with speculative editing.
Neural Search Space in Gboard Decoder (2024.emnlp-industry)

Copied to clipboard

Challenge: Gboard decoder uses context, a lexicon and language models to provide a user-friendly keyboard.
Approach: They propose a Neural Search Space which replaces an N-gram LM with a neural network LM and dynamically constructs the search space during decoding.
Outcome: The proposed system improves the quality of the decoded keyboards on various locales with acceptable latency increases.
TreeReview: A Dynamic Tree of Questions Framework for Deep and Efficient LLM-based Scientific Peer Review (2025.emnlp-main)

Copied to clipboard

Challenge: Large Language Models (LLMs) have shown significant potential in assisting peer review, but current methods struggle to generate thorough and insightful reviews while maintaining efficiency.
Approach: They propose a framework that models paper review as a hierarchical and bidirectional question-answering process.
Outcome: The proposed framework outperforms baselines on full review generation and actionable feedback comments generation tasks while reducing LLM token usage by up to 80% compared to computationally intensive approaches.
Specialization through Collaboration: Understanding Expert Interaction in Mixture-of-Expert Large Language Models (2026.eacl-long)

Copied to clipboard

Challenge: Mixture-of-Experts (MoE) based large language models are popular for multitasking . however, whether each expert can specialize to a task remains unclear .
Approach: They propose to use a dictionary learning approach to analyze expert collaboration mechanisms in MoE LLMs.
Outcome: The proposed model outperforms existing methods by 2.5% while enabling 50% expert reduction.
Beyond Surface Alignment: Rebuilding LLMs Safety Mechanism via Probabilistically Ablating Refusal Direction (2025.findings-emnlp)

Copied to clipboard

Challenge: Jailbreak attacks pose persistent threats to large language models . current safety alignment methods have insufficient safety alignment depth and unrobust internal defense mechanisms.
Approach: a new safety alignment framework is developed to overcome jailbreak attacks . the framework forces the model to dynamically rebuild its refusal mechanisms from jailbreak states .
Outcome: a new safety alignment framework reduces attack success rates by approximately 95% on four open-source LLM families and six representative attacks.
Detecting RAG Extraction Attack via Dual-Path Runtime Integrity Game (2026.acl-long)

Copied to clipboard

Challenge: Retrieval-Augmented Generation (RAG) systems augment large language models with external knowledge, but introduce a critical security vulnerability: Knowledge Base Leakage.
Approach: They propose a runtime defense mechanism inspired by stack canaries in software security . canaryRAG embeds carefully designed canary tokens into retrieved chunks and reformulates RAG extraction defense as a dual-path runtime integrity game.
Outcome: The proposed system can detect and prevent RAG Knowledge Base Leakage in real time . it can be integrated into arbitrary RAG pipelines without retraining or structural modifications .
Proofread: Fixes All Errors with One Tap (2024.acl-demos)

Copied to clipboard

Challenge: Extensive experiments on a human-labeled golden set showed our tuned PaLM2-XS model achieved 85.56% good ratio.
Approach: They propose a two-stage tuning approach to acquire the dedicated Large Language Model for the feature, followed by a reinforcement learning approach for targeted refinement.
Outcome: The proposed model achieves 85.56% good quality on Rewrite and proofread tasks on human-labeled golden sets.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations