Papers by Zhicheng Zheng

21 papers
RAG-Studio: Towards In-Domain Adaptation of Retrieval Augmented Generation Through Self-Alignment (2024.findings-emnlp)

Copied to clipboard

Challenge: Existing RAG systems that use pre-trained LLMs and retrievers often fail in specialized domains and applications.
Approach: They propose a self-aligned training framework that adapts general RAG models to specific domains solely through synthetic data.
Outcome: Experiments on specialized domain corpus, general LLM, and general retriever show that the self-aligned training framework outperforms human-annotated training data in specialized fields.
GMSA: Enhancing Context Compression via Group Merging and Layer Semantic Alignment (2026.acl-long)

Copied to clipboard

Challenge: Large Language Models (LLMs) have achieved remarkable performance across NLP tasks . however, in long-context scenarios, they face high computational cost and information redundancy.
Approach: They propose an encoder-decoder context compression framework that generates a compact sequence of soft tokens for downstream tasks.
Outcome: Experiments show that GMSA outperforms baselines on multiple long-context question answering and summarization benchmarks while maintaining low end-to-end latency.
Perception Compressor: A Training-Free Prompt Compression Framework in Long Context Scenarios (2025.findings-naacl)

Copied to clipboard

Challenge: Long prompts contain redundant information and are sensitive to the position of key information in long context scenarios.
Approach: They propose a training-free prompt compression framework that retains key information at token level while removing distracting tokens.
Outcome: The proposed framework outperforms existing methods on long context benchmarks.
Flaming-hot Initiation with Regular Execution Sampling for Large Language Models (2025.findings-naacl)

Copied to clipboard

Challenge: Large language models (LLMs) have demonstrated remarkable capabilities across various domains since the release of ChatGPT . a key challenge in developing these general capabilities is efficiently sourcing diverse, high-quality data.
Approach: They introduce Flaming-hot Initiation with Regular Execution (FIRE) sampling to efficiently find good responses by promoting diversity.
Outcome: The proposed method enhances inference-time generation quality and benefits training in the alignment stage.
InquireMobile: Teaching VLM-based Mobile Agent to Request Human Assistance via Reinforcement Fine-Tuning (2026.acl-long)

Copied to clipboard

Challenge: Recent advances in Vision-Language Models (VLMs) have enabled mobile agents to perceive and interact with real-world mobile environments based on human instructions.
Approach: They propose a vision-language model that actively seeks human confirmation at critical decision points and a model inspired by reinforcement learning.
Outcome: The proposed model achieves an improvement of 46.8% in inquiry success rate and the best overall success rate among existing baselines on InquireBench.
A Multi-Task Embedder For Retrieval Augmented LLMs (2024.acl-long)

Copied to clipboard

Challenge: Existing retrieval methods face limitations in terms of knowledge, memory, and action.
Approach: They propose a retrieval enhancement mechanism that brings in useful information from external sources to augment the LLM.
Outcome: The proposed method significantly improves the LLM’s performance in various downstream tasks while introducing superior retrieval augmentation’s effect over both general and task-specifc retrievers.
INTERS: Unlocking the Power of Large Language Models in Search with Instruction Tuning (2024.acl-long)

Copied to clipboard

Challenge: Large language models (LLMs) have demonstrated impressive capabilities in various natural language processing tasks, but their application to information retrieval tasks is still challenging due to the infrequent occurrence of many IR-specific concepts in natural language.
Approach: They propose to use instruction tuning to enhance LLMs' proficiency in IR tasks by combining a dataset with manually written templates to analyze the effects of instruction design, template diversity, few-shot demonstrations, and the volume of instructions.
Outcome: The proposed model can be used to perform query understanding, document understanding, and query-document relationship understanding tasks.
ChatRetriever: Adapting Large Language Models for Generalized and Robust Conversational Dense Retrieval (2024.emnlp-main)

Copied to clipboard

Challenge: a conversational search system requires accurate interpretation of user intent from complex multi-turn contexts.
Approach: They propose a dual-learning approach that adapts LLMs for retrieval via contrastive learning while enhancing the complex session understanding through masked instruction tuning.
Outcome: The proposed approach outperforms existing retrieval methods on five conversational search benchmarks.
Read As Human: Compressing Context via Parallelizable Close Reading and Skimming (2026.acl-long)

Copied to clipboard

Challenge: Existing task-aware methods require loading the entire input sequence at once for compression, which suffer from computational inefficiency.
Approach: They propose a framework that adopts an adaptive hybrid reading strategy to reduce computational inefficiency and redundant information in long-context scenarios.
Outcome: Experiments show that RAM outperforms baselines on multiple question answering and summarization benchmarks while delivering up to a 12x speedup on long inputs.
FineRAG: Fine-grained Retrieval-Augmented Text-to-Image Generation (2025.coling-main)

Copied to clipboard

Challenge: Recent advances in text-to-image generation still exhibit limitations in terms of knowledge access.
Approach: They propose a fine-grained retrieval-augmented image generation model that breaks down the retrieval task into four critical stages: query decomposition, candidate selection, retrieval augmented diffusion, and self-reflection.
Outcome: The proposed method significantly reduces noise associated with retrieval-augmented image generation and performs better in complex, open-world scenarios.
Grounding Language Model with Chunking-Free In-Context Retrieval (2024.acl-long)

Copied to clipboard

Challenge: CFIC retrieval approach eliminates the need for document chunking and provides a more efficient and efficient method for RAG systems.
Approach: They propose a Chunking-Free In-Context retrieval approach specifically tailored for RAG systems . they employ auto-aggressive decoding to accurately identify specific evidence text .
Outcome: The proposed method is better than traditional methods on open question answering datasets.
How Credible Is an Answer From Retrieval-Augmented LLMs? Investigation and Evaluation With Multi-Hop QA (2025.coling-main)

Copied to clipboard

Challenge: Retrieval-augmented large language models (RaLLMs) are reshaping knowledge acquisition, offering long-form, knowledge-grounded answers through advanced reasoning and generation capabilities.
Approach: They propose a benchmarking system to evaluate RaLLMs' correctness and Groundedness to determine their reliability in multi-hop question-answering tasks.
Outcome: The proposed model-based evaluation pipeline for multi-hop question-answering tasks reveals that the model generates inaccuracies when dealing with flawed or partial knowledge.
Single LLM, Multiple Roles: A Unified Retrieval-Augmented Generation Framework Using Role-Specific Token Optimization (2025.emnlp-main)

Copied to clipboard

Challenge: Existing studies have optimized retrieval-augmented generation (RAG) across sub-tasks, but integrating these optimizations into a unified framework remains challenging.
Approach: They propose a unified retrieval-augmented generation framework that optimizes role-specific tokens for multi-task processing.
Outcome: The proposed framework achieves efficient multi-task processing through role-specific token optimization.
Coarse-to-Fine: Hierarchical Multi-task Learning for Natural Language Understanding (2022.coling-1)

Copied to clipboard

Challenge: Existing methods to learn downstream tasks by stitches skill block lack rationality and interpretation.
Approach: They propose a hierarchical framework with a coarse-to-fine paradigm for generalized text representations from the large-scale corpus.
Outcome: The proposed model learns basic language properties from all tasks and boosts performance on relevant tasks.
Boosting Long-Context Information Seeking via Query-Guided Activation Refilling (2025.acl-long)

Copied to clipboard

Challenge: Existing methods for processing long contexts are ineffective due to their inherent context window limitations and the computational burden of extensive key-value activations.
Approach: They propose a method for processing long context information-seeking tasks via query-guided ACtivation REfilling (ACRE) a bi-layer KV Cache is constructed where the layer-1 cache compactly captures global information and the layer-2 cache provides detailed, localized information.
Outcome: The proposed method achieves significant improvements in both performance and efficiency on a variety of long-context information-seeking datasets.
Lightweight LLM Agent Memory with Small Language Models (2026.acl-long)

Copied to clipboard

Challenge: Existing external memory systems for LLMs have low online overhead but are unstable in accumulating latency over long interactions.
Approach: They propose a lightweight memory system for better agent memory driven by Small Language Models . lightmem modularizes memory retrieval, writing, and long-term consolidation . they show consistent gains across model scales and high efficiency .
Outcome: The proposed system improves agent memory but has low latency and low online overhead . it separates online processing from offline consolidation to enable efficient memory invocation . the proposed system achieves an average F1 improvement of 2.5 over A-MEM on LoCoMo .
Hybrid Inverted Index Is a Robust Accelerator for Dense Retrieval (2023.emnlp-main)

Copied to clipboard

Challenge: Inverted file structure is a common technique for accelerating dense retrieval, but its lossy nature degrades it.
Approach: They propose a hybrid index where embedding clusters and salient terms work collaboratively to accelerate dense retrieval.
Outcome: The proposed method achieves lossless retrieval quality with competitive efficiency across index settings.
Chinese SimpleQA: A Chinese Factuality Evaluation for Large Language Models (2025.acl-long)

Copied to clipboard

Challenge: Current frontier models sometimes generate false outputs or answers that are not substantiated by evidence.
Approach: They propose Chinese SimpleQA, a Chinese benchmark to evaluate LLMs' factuality . they focus on Chinese language over 6 major topics with 99 diverse subtopics .
Outcome: The Chinese SimpleQA benchmark evaluates the factuality ability of LLMs . the questions and answers are short and easy-to-evaluate .
VC4VG: Optimizing Video Captions for Text-to-Video Generation (2025.emnlp-main)

Copied to clipboard

Challenge: Recent advances in text-to-video generation highlight the critical role of high-quality video-text pairs in training models capable of producing coherent and instruction-aligned videos.
Approach: They propose a caption optimization framework tailored to the needs of T2V models.
Outcome: The proposed framework improves video caption quality and video generation performance.
Can Large Language Models Detect Errors in Long Chain-of-Thought Reasoning? (2025.acl-long)

Copied to clipboard

Challenge: Recent advances in o1-like models have generated long Chain-of-Thought reasoning steps to improve the reasoning abilities of existing Large Language Models (LLMs).
Approach: They propose a DeltaBench to analyze the quality and effectiveness of o1-like models and measure their ability to detect errors in long COT reasoning.
Outcome: The proposed model can detect errors in long COT reasoning.
PaperScope: A Multi-Modal Multi-Document Benchmark for Agentic Deep Research Across Massive Scientific Papers (2026.findings-acl)

Copied to clipboard

Challenge: Existing benchmarks focus on single-document understanding, whereas real scientific workflows require integrating evidence from multiple papers.
Approach: They propose a multi-modal multi-document benchmark for agentic deep research that integrates evidence from multiple documents.
Outcome: Experimental results show that even advanced systems achieve limited scores on PaperScope . paper provides a rigorous benchmark alongside a pipeline for constructing large multi-modal, multi-source deep research datasets.

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