Papers by Jian Luan

30 papers
Mixture of Diverse Size Experts (2024.emnlp-industry)

Copied to clipboard

Challenge: Recent large language models (LLMs) have shown superior performance in a variety of tasks due to the sub-linearly increasing computational costs.
Approach: They propose a new MoE architecture with designed layers where experts have different sizes to mitigate this defect.
Outcome: The proposed architecture surpasses existing MoEs by adaptively assigning the parameter budget to experts while maintaining the same total parameter size and number of experts.
MAKAR: a Multi-Agent framework based Knowledge-Augmented Reasoning for Grounded Multimodal Named Entity Recognition (2025.emnlp-main)

Copied to clipboard

Challenge: Existing methods for GMNER fail to address semantic ambiguity caused by polysemy and long-tail distribution of datasets.
Approach: They propose a framework for Grounded Multimodal Named Entity Recognition that leverages a Multimodal Large Language Model to address semantic ambiguity.
Outcome: Extensive experiments show that the proposed framework outperforms existing methods on two benchmark datasets.
Demystifying Small Language Models for Edge Deployment (2025.acl-long)

Copied to clipboard

Challenge: Small language models (SLMs) are a promising solution for resource-constrained devices such as smartphones and the Web of Things.
Approach: They propose to use SLMs to build and optimize a set of small language models that are publicly accessible.
Outcome: The proposed models outperform 7B models in general tasks, while their in-context learning capabilities remain limited and their efficiency has significant optimization potential.
Pruning Large Language Models to Intra-module Low-rank Architecture with Transitional Activations (2024.findings-acl)

Copied to clipboard

Challenge: Structured pruning is a feasible solution for end-side LLM deployment . however, achieving a high compression ratio for scaled-up LLMs remains a challenge .
Approach: They propose a task-agnostic structured pruning approach coupled with a compact Transformer architecture to prune LLMs into an intra-module low-rank architecture.
Outcome: The proposed approach reduces transitional activations inside multi-head attention (MHA) and multi-layer perceptron (MLP) modules while preserving inter-module activations sensitive to perturbations.
BacktrackAgent: Enhancing GUI Agent with Error Detection and Backtracking Mechanism (2025.emnlp-main)

Copied to clipboard

Challenge: Existing GUI agents focus on enhancing the accuracy of individual actions and lack effective mechanisms for detecting and recovering from errors.
Approach: They propose a framework that incorporates a backtracking mechanism to enhance the task completion capabilities of GUI agents by verifier, judger, and reflector components.
Outcome: The proposed framework improves task success rate and step accuracy on Mobile3M and Auto-UI benchmarks.
A Comprehensive Evaluation of Quantization Strategies for Large Language Models (2024.findings-acl)

Copied to clipboard

Challenge: Quantization studies have focused on instruction-tuned LLMs, leaving their performance on other benchmarks unclear.
Approach: They propose a framework to evaluate quantized large language models using four dimensions . they propose to reduce the bits needed for model weights or activations with minimal performance loss .
Outcome: The proposed framework can retain comparable performance to non-quantized LLMs on most benchmarks.
Mobile-Bench: An Evaluation Benchmark for LLM-based Mobile Agents (2024.acl-long)

Copied to clipboard

Challenge: Existing benchmarks for LLM-based mobile agents are insufficient to evaluate their capabilities.
Approach: They propose a benchmark to evaluate LLM-based mobile agents' planning capabilities . they expand UI operations by incorporating 103 APIs to accelerate task completion .
Outcome: The proposed benchmarks are based on 103 collected APIs and real user queries . the data is categorized into three distinct groups: SAST, SAMT, and MAMT .
Global Eye: Breaking the “Fixed Thinking Pattern” during the Instruction Expansion Process (2025.acl-long)

Copied to clipboard

Challenge: Existing methods focus on constructing multi-perspective prompts to expand instructions, overlooking the “Fixed Thinking Pattern” issue of Large Language Models.
Approach: They propose a method that analyzes the statistical characteristics of newly generated instructions and updates the prompts after a fixed number of instruction expansions.
Outcome: The proposed method surpasses open-source LLMs and GPT3.5 in several metrics.
Multilingual Machine Translation with Open Large Language Models at Practical Scale: An Empirical Study (2025.naacl-long)

Copied to clipboard

Challenge: Large language models (LLMs) have shown continuously improving multilingual capabilities.
Approach: They evaluate the ability of open LLMs to handle multilingual machine translation tasks using a parallel-first monolingual-second data mixing strategy.
Outcome: The proposed model outperforms state-of-the-art models and achieves competitive performance with Google Translate and GPT-4-turbo.
TailorKV: A Hybrid Framework for Long-Context Inference via Tailored KV Cache Optimization (2025.findings-acl)

Copied to clipboard

Challenge: Existing work mitigates memory overhead by offloading or compressing the Key-Value cache.
Approach: They propose a method that integrates quantization and offloading into a generative large language model by using a hybrid compression method.
Outcome: The proposed method outperforms the state-of-the-art in long-context evaluations.
DetermLR: Augmenting LLM-based Logical Reasoning from Indeterminacy to Determinacy (2024.acl-long)

Copied to clipboard

Challenge: Recent advances in large language models (LLMs) have revolutionized the landscape of reasoning tasks.
Approach: They propose a new approach that rethinks the reasoning process as an evolution from indeterminacy to determinacy.
Outcome: The proposed model surpasses all baselines on various logical reasoning benchmarks.
Revisiting Entropy in Reinforcement Learning for Large Reasoning Models (2026.findings-acl)

Copied to clipboard

Challenge: Reinforcement learning with verifiable rewards (RLVR) has emerged as a paradigm for enhancing the reasoning capabilities of large language models.
Approach: They propose a positive-advantage reweighting approach that regulates model entropy by adjusting the loss weights assigned to tokens with positive advantages during RLVR training.
Outcome: The proposed approach regulates model entropy by adjusting loss weights assigned to tokens with positive advantages during RLVR training while maintaining competitive performance.
PMSS: Pretrained Matrices Skeleton Selection for LLM Fine-tuning (2025.coling-main)

Copied to clipboard

Challenge: Low-rank adaptation and its variants have been popular due to their ability to avoid excessive inference costs.
Approach: They propose a low-rank adaptation method that enables high-rank updates with low costs while leveraging semantic and linguistic information inherent in pre-trained weight.
Outcome: The proposed method outperforms LoRA and other fine-tuning methods across tasks with less trainable parameters.
ReachAgent: Enhancing Mobile Agent via Page Reaching and Operation (2025.naacl-long)

Copied to clipboard

Challenge: Existing mobile AI agents focus on most task-relevant elements at each step, leading to local optimal solutions and ignoring the overall GUI flow.
Approach: They propose a mobile AI agent that breaks tasks into page reaching and operation subtasks and a framework that focuses on improving its task-completion abilities.
Outcome: The proposed framework improves IoU accuracy and text accuracy by 7.12% and 7.69% on step-level and 4.72% and 4.63% on task-level compared to the SOTA agent.
MobileVLM: A Vision-Language Model for Better Intra- and Inter-UI Understanding (2024.findings-emnlp)

Copied to clipboard

Challenge: Recent mobile AI agents based on VLMs lack basic mobile capabilities due to their pre-trained nature.
Approach: They propose a mobile AI agent based on VLMs that includes additional pre-training stages to enhance both intra- and inter-UI understanding.
Outcome: The proposed model outperforms existing VLMs on the Chinese mobile dataset Mobile3M .
End-to-End Optimization of LLM-Driven Multi-Agent Search Systems via Heterogeneous-Group-Based Reinforcement Learning (2026.acl-long)

Copied to clipboard

Challenge: Existing multi-agent reinforcement learning methods depend on large critic networks to evaluate joint actions, leading to instability and high memory costs.
Approach: They propose a method to optimize large language models for agent-specific roles . they propose combining agent-based frameworks with retrieval-augmented generation .
Outcome: Experiments show that multi-agent group policy optimization outperforms baselines in task performance and computational efficiency.
Exploring All-In-One Knowledge Distillation Framework for Neural Machine Translation (2023.emnlp-main)

Copied to clipboard

Challenge: Existing knowledge distillation methods only obtain one lightweight student each time . this could be resource-intensive and resulting in multiple students not being optimally utilized .
Approach: They propose a knowledge distillation framework which generates multiple satisfactory students at once.
Outcome: The proposed framework generates multiple satisfactory students at once.
STEP: Success-Rate-Aware Trajectory-Efficient Policy Optimization (2026.findings-acl)

Copied to clipboard

Challenge: Existing GRPO-based methods allocate sampling uniformly across tasks regardless of difficulty, propagate misleading learning signals and incur high sample-collection costs.
Approach: They propose a framework that allocates sampling based on per-task success rates and performs fine-grained step-level optimization.
Outcome: The proposed method improves sample efficiency and training stability over existing GRPO variants and three ablation variants on OSWorld and AndroidWorld.
Weaving Context Across Images: Improving Vision-Language Models through Focus-Centric Visual Chains (2025.acl-long)

Copied to clipboard

Challenge: Existing vision-language models struggle to disentangle information scattered across complex visual inputs, leading to performance degradation.
Approach: They propose a focus-centric visual chain paradigm that enhances VLMs’ perception, comprehension, and reasoning abilities in multi-image scenarios.
Outcome: The proposed approach achieves average performance gains of 3.16% and 2.24% across two distinct model architectures, without compromising the general vision-language capabilities.
SPO: Self Preference Optimization with Self Regularization (2025.findings-emnlp)

Copied to clipboard

Challenge: Existing reference-free preference optimization methods exhibit higher training efficiency but are prone to overoptimization, leading to performance degradation.
Approach: They propose a reference-free preference optimization method that replaces the logsigmoid loss function with a SiLU function to improve the model's performance.
Outcome: The proposed method achieves 7% improvement over SimPO on AlpacaEval 2 and MT-Bench.
Attention Basin: Why Contextual Position Matters in Large Language Models (2026.acl-long)

Copied to clipboard

Challenge: Large Language Models (LLMs) are sensitive to the contextual position of information in input.
Approach: They introduce Attention-Driven Reranking (AttnRank) which estimates a model’s intrinsic positional attention preferences using a small calibration set and reorders retrieved documents or few-shot examples to align the most salient content with these high-attention positions.
Outcome: Experiments on multi-hop QA and few-shot in-context learning tasks show that AttnRank achieves substantial improvements across 10 large language models of varying architectures and scales, without modifying model parameters or training procedures.
ToolPlanner: A Tool Augmented LLM for Multi Granularity Instructions with Path Planning and Feedback (2024.emnlp-main)

Copied to clipboard

Challenge: Recent studies have shown that tool-augmented large language models can interact with external tools in multiple rounds and provide a final answer.
Approach: They propose a tool-augmented large language model that can interact with external tools in multiple rounds and provide a final answer to an instruction.
Outcome: The proposed framework significantly improves Match Rate, Pass Rate and Win Rate by 26.8%, 20.2%, and 5.6% compared to the SOTA model.
HoPE: A Novel Positional Encoding Without Long-Term Decay for Enhanced Context Awareness and Extrapolation (2025.acl-long)

Copied to clipboard

Challenge: Existing positional encodings exhibit long-term decay, based on an entrenched and long-standing opinion that tokens farther away from the current position carry less relevant information.
Approach: They propose a high-frequency rotary position encoding (HoPE) that replaces specific components in RoPE with position-independent ones, retaining only high- frequency signals.
Outcome: The proposed method exhibits greater robustness to the out-of-distribution behavior in attention patterns during extrapolation.
Exploring Better Text Image Translation with Multimodal Codebook (2023.acl-long)

Copied to clipboard

Challenge: Current studies on text image translation face bottlenecks due to lack of a publicly available dataset and poor optical character recognition.
Approach: They propose a text image translation model with a multimodal codebook and an OCR dataset for Chinese-English translation.
Outcome: The proposed model can associate the image with relevant texts, providing useful supplementary information for translation.
Doc-V*: Coarse-to-Fine Interactive Visual Reasoning for Multi-Page Document VQA (2026.acl-long)

Copied to clipboard

Challenge: Existing OCR-free approaches to document visual question answering are brittle and passive.
Approach: They propose an OCR-free agentic framework that casts multi-page DocVQA as sequential evidence aggregation.
Outcome: The proposed framework outperforms open-source and proprietary models in five benchmarks and improves out-of-domain performance by 47.9% over baseline.
VecInfer: Efficient LLM Inference with Low-Bit KV Cache via Outlier-Suppressed Vector Quantization (2026.acl-long)

Copied to clipboard

Challenge: Existing quantization methods for large language models suffer performance degradation at ultra-low bit-widths due to key cache outliers.
Approach: They propose a vector quantization method that suppresses outliers in the key cache and reduces memory access overhead.
Outcome: The proposed method outperforms baseline quantization methods across long-context understanding and mathematical reasoning tasks while minimizing memory access overhead.
ToolRerank: Adaptive and Hierarchy-Aware Reranking for Tool Retrieval (2024.lrec-main)

Copied to clipboard

Challenge: Recent studies have proposed tool learning, which augments LLMs with external tools.
Approach: They propose an adaptive and hierarchy-aware reranking method to refine retrieval results by truncating the retrieval result related to seen and unseen tools at different positions.
Outcome: The proposed method improves retrieval results, leading to better execution results generated by the LLM.
More is not always better? Enhancing Many-Shot In-Context Learning with Differentiated and Reweighting Objectives (2025.acl-long)

Copied to clipboard

Challenge: Large language models excel at few-shot in-context learning but performance plateaus as ICL demonstrations increase from a few to many.
Approach: They propose a novel optimization method that optimizes the negative log-likelihood objective and reweights the model to achieve many-shot performance.
Outcome: The proposed method achieves significant performance improvements across a large-scale dataset.
Browsing Like Human: A Multimodal Web Agent with Experiential Fast-and-Slow Thinking (2025.acl-long)

Copied to clipboard

Challenge: Existing web agents lack visual perception, planning, and memory abilities, but their reasoning process is deviate from human cognition.
Approach: They propose a multimodal web agent framework that emulates human planning process to decompose complex user instructions.
Outcome: The proposed framework emulates human planning process to decompose complex user instructions.
MobileBench-OL: A Comprehensive Chinese Benchmark for Evaluating Mobile GUI Agents in Real-World Environment (2026.findings-acl)

Copied to clipboard

Challenge: Recent advances in mobile Graphical User Interface (GUI) agents highlight the growing need for comprehensive evaluation benchmarks.
Approach: They propose an online benchmark with 1080 tasks from 80 Chinese apps that measures task execution, complex reasoning, noise robustness and auto-eval framework with a reset mechanism.
Outcome: The proposed benchmark measures task execution, complex reasoning, and noise robustness of agents by including 5 subsets, which set multiple evaluation dimensions.

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