Papers by Xiaodong Han

10 papers
Understanding the Difficulty of Training Transformers (2020.emnlp-main)

Copied to clipboard

Challenge: Admin (Adaptive model initialization) is more stable, converges faster, and leads to better performance.
Approach: They propose a model initialization algorithm to stabilize early training and unleash its full potential in the late stage.
Outcome: The proposed model initialization method stabilizes early training and unleashes full potential in late stage.
Stabilizing Efficient Reasoning with Step-Level Advantage Selection (2026.findings-acl)

Copied to clipboard

Challenge: Large language models generate long and verbose reasoning traces at inference time . short context post-training alone induces substantial reasoning compression .
Approach: They propose a step-level advantage selection approach that reduces reasoning length by over 30% . they propose to use GRPO without any length-aware objective to train models in a shorter context window .
Outcome: The proposed approach reduces average reasoning length by over 30% while improving Pass@1 accuracy by 3.79 points over the strongest length-aware baseline.
PyramidInfer: Pyramid KV Cache Compression for High-throughput LLM Inference (2024.findings-acl)

Copied to clipboard

Challenge: Existing methods to reduce memory usage for large language models neglect inter-layer dependency between layers and huge memory consumption in pre-computation.
Approach: They propose a method that compresses the KV cache by layer-wise retaining crucial context.
Outcome: The proposed method reduces memory usage by layer-wise retaining crucial context . it can improve 2.2x throughput compared to Accelerate with over 54% memory reduction .
AUGUST: an Automatic Generation Understudy for Synthesizing Conversational Recommendation Datasets (2023.findings-acl)

Copied to clipboard

Challenge: Existing work on conversational recommendation systems lacks high-quality data . existing datasets lack large-scale and high-level data based on human annotators .
Approach: They propose an automatic dataset synthesis approach that generates large-scale recommendation dialogues using structured graphs based on user-item information from the real world.
Outcome: The proposed approach can generate large-scale and high-quality recommendation dialogues . it exploits user preferences, knowledge graphs, and conversation ability from existing datasets based on real-world data .
RESIN: A Dockerized Schema-Guided Cross-document Cross-lingual Cross-media Information Extraction and Event Tracking System (2021.naacl-demos)

Copied to clipboard

Challenge: We present a new information extraction system that can construct temporal event graphs from news documents.
Approach: They propose a temporal event graph extraction system that can extract news documents . they extend the system from sentence-level event extraction to cross-document cross-media event extraction .
Outcome: The proposed system can extract temporal event graphs from news documents in multiple languages and multiple data modalities.
MatPlotAgent: Method and Evaluation for LLM-Based Agentic Scientific Data Visualization (2024.findings-acl)

Copied to clipboard

Challenge: Scientific data visualization is an essential process in research, but its use of large language models remains unexplored.
Approach: They propose a model-agnostic LLM agent framework to automate scientific data visualization tasks.
Outcome: The proposed framework improves performance of commercial and open-source models.
The Devil in Linear Transformer (2022.emnlp-main)

Copied to clipboard

Challenge: Existing linear transformers suffer from performance degradations on various tasks and corpus.
Approach: They propose a new linear attention that replaces scaling with a normalization to stabilize gradients and confine attention to neighbouring tokens in early layers.
Outcome: The proposed model outperforms vanilla transformers on the long-range arena benchmark while being significantly more space-time efficient.
Generation-Augmented Retrieval for Open-Domain Question Answering (2021.acl-long)

Copied to clipboard

Challenge: Existing approaches to answer open-domain questions use sparse representations and sparsity.
Approach: They propose a method which augments a query by generating relevant contexts from heuristically discovered contexts without external supervision.
Outcome: The proposed approach outperforms state-of-the-art dense retrieval methods on natural questions and triviaQA datasets.
LLM×MapReduce: Simplified Long-Sequence Processing using Large Language Models (2025.acl-long)

Copied to clipboard

Challenge: Existing studies have focused on extending the context length of large language models (LLMs) due to their quadratic computational complexity and a lack of high-quality long training examples, most LLMs are trained with a limited window size.
Approach: They propose a training-free framework that enables large language models to effectively process long texts using a divide-and-conquer strategy for comprehensive document understanding.
Outcome: The proposed framework outperforms open-source and commercial long-context LLMs and is compatible with several models.
Reader-Guided Passage Reranking for Open-Domain Question Answering (2021.findings-acl)

Copied to clipboard

Challenge: Current open-domain question answering systems follow a Retriever-Reader architecture . current systems do not use a reranker, which reranked passages based on top predictions of the reader .
Approach: They propose a reader-guIDEd reranking method that reranked passages based on top predictions . they show that RIDER achieves 10 to 20 absolute gains in top-1 retrieval accuracy .
Outcome: The proposed method achieves 10 to 20 gains in top-1 retrieval accuracy and 1 to 4 Exact Match gains without training.

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