Papers by Peijie Huang
Too Long, Do Re-weighting for Efficient LLM Reasoning Compression (2026.acl-long)
Copied to clipboard
Zhong-Zhi Li, Xiao Liang, Zihao Tang, Lei Ji, Peijie Wang, Haotian Xu, Xing W, Haizhen Huang, Weiwei Deng, Yeyun Gong, Zhijiang Guo, Xiao Liu, Fei Yin, Cheng-Lin Liu
| Challenge: | Large Language Models (LLMs) have recently achieved remarkable progress on complex reasoning tasks by leveraging extended Chain-of-Thought (CoT) techniques. |
| Approach: | They propose a method that uses Extended Chain-of-Thought (EFT) to reduce the number of output tokens by nearly 40% while maintaining the accuracy of the reasoning. |
| Outcome: | The proposed method reduces the number of output tokens by nearly 40% while maintaining the accuracy of the reasoning. |
From Noise to Clarity: Filtering Real and LLM-Generated Samples for Enhanced Intent Detection (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Using large language models for generating synthetic samples for data augmentation can cause problems with the generalization ability of classification models. |
| Approach: | They propose an interpretable Sample Filter by Topic Modeling framework that allows for filtering by topic and a 'sampler by topic' framework. |
| Outcome: | The proposed framework reduces the quantity of real and synthetic samples while improving the performance of the classification models. |
ECLM: Entity Level Language Model for Spoken Language Understanding with Chain of Intent (2025.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have demonstrated impressive capabilities in language generation and general task performance, but their application to spoken language understanding remains challenging. |
| Approach: | They propose an Entity-level Language Model framework which reformulates slot-filling as an entity recognition task and introduces a new concept, Chain of Intent, to enable step-by-step multi-intent recognition. |
| Outcome: | The proposed framework outperforms strong baselines such as Uni-MIS and achieves gains of 3.7% and 3.1% on MixATIS and MixSNIPS. |
DMIN: A Discourse-specific Multi-granularity Integration Network for Conversational Aspect-based Sentiment Quadruple Analysis (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing studies focus on enhancing token-level interactions, but lack sufficient modeling of discourse structure information. |
| Approach: | They propose to use a discourse structure called "thread" to enhance token interaction among different utterances. |
| Outcome: | The proposed model achieves state-of-the-art on two datasets. |
MIDLM: Multi-Intent Detection with Bidirectional Large Language Models (2025.coling-main)
Copied to clipboard
| Challenge: | Existing models that use autoregressive architectures restrict the sharing of token information within a sentence. |
| Approach: | They propose a framework that integrates intent number detection and multi-intent selection to enable autoregressive LLMs to leverage bidirectional information awareness through post-training. |
| Outcome: | The proposed framework outperforms existing models and pretrained baselines in the multi-intent detection task. |
Logits Reranking via Semantic Labels for Hard Samples in Text Classification (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Existing research on text classification models ignores the semantic information inherent in labels, treating them as one-hot vectors. |
| Approach: | They propose a model-agnostic method that leverages label semantics and auto detection of hard samples to improve classification accuracy. |
| Outcome: | The proposed method shows significant improvements across different PLMs. |
Synergistic Augmentation: Enhancing Cross-Domain Zero-Shot Slot Filling with Small Model-Assisted Large Language Models (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing approaches to slot filling are limited due to data scarcity and timeconsuming efforts. |
| Approach: | They propose a framework that harnesses the power of a small model to augment inferential capabilities of LLMs without additional training. |
| Outcome: | The proposed framework improves slot filling performance on a spoken language dataset and a NER dataset. |