Papers by Ziwei Liu

30 papers
Dynamic Parallel Tree Search for Efficient LLM Reasoning (2025.acl-long)

Copied to clipboard

Challenge: Recent methods focus on search accuracy while overlooking computational efficiency.
Approach: They propose a parallelism framework that dynamically optimizes reasoning path in inference.
Outcome: The proposed framework improves efficiency by 2-4 on average while maintaining or even surpassing existing reasoning algorithms in accuracy.
Pragmatics in the Era of Large Language Models: A Survey on Datasets, Evaluation, Opportunities and Challenges (2025.acl-long)

Copied to clipboard

Challenge: linguistics studies how context influences meaning of language and how people use it to convey implied meanings, emotions, and intentions.
Approach: They analyze task designs, data collection methods, evaluation approaches and their relevance to real-world applications.
Outcome: The findings highlight emerging trends, challenges, and gaps in existing benchmarks . the findings will contribute to more nuanced and context-aware NLP models .
Denoising Bottleneck with Mutual Information Maximization for Video Multimodal Fusion (2023.acl-long)

Copied to clipboard

Challenge: Prior denoising methods suppress redundant and noisy information at risk of losing critical information.
Approach: They propose a denoising bottleneck fusion model for fine-grained video multimodal fusion . they employ a bottleneck mechanism to filter out noise and redundancy with a restrained receptive field .
Outcome: The proposed model improves on state-of-the-art video multimodal fusion benchmarks.
ImageNetVC: Zero- and Few-Shot Visual Commonsense Evaluation on 1000 ImageNet Categories (2023.findings-emnlp)

Copied to clipboard

Challenge: Large Language Models (LLMs) are becoming general-purpose APIs, requiring visual knowledge to be understood.
Approach: They propose to evaluate the visual capability of large-scale large-language models through visual commonsense evaluation using a human-annotated dataset.
Outcome: The proposed dataset compares the visual commonsense knowledge of large-scale models with those of unimodal LLMs and visually augmented models.
Video-MMMU: Evaluating Knowledge Acquisition from Multidisciplinary Professional Videos (2026.acl-long)

Copied to clipboard

Challenge: Existing video benchmarks do not evaluate the knowledge acquisition capabilities of Large Multimodal Models (LMMs) existing video benchmark focuses on static, general visual understanding tasks, without evaluating whether models can acquire knowledge dynamically.
Approach: They propose a multi-modal, multi-discipline, multitrack benchmark that evaluates Large Multimodal Models’ ability to acquire knowledge from college-level, educational videos.
Outcome: The proposed benchmark reveals a substantial gap between human learners and current Large Multimodal Models (LMMs) and focuses on improving their learning efficiency.
HiDe-LLaVA: Hierarchical Decoupling for Continual Instruction Tuning of Multimodal Large Language Model (2025.acl-long)

Copied to clipboard

Challenge: Existing methods to improve instructionfollowing performance of MLLMs often trade off memory efficiency for performance gains, compromising overall efficiency.
Approach: They propose a task-specific expansion and task-general fusion framework based on variations in Centered Kernel Alignment (CKA) similarity across different model layers when trained on diverse datasets.
Outcome: The proposed framework improves performance compared to existing benchmarks.
Trajectory2Task: Training Robust Tool-Calling Agents with Synthesized Yet Verifiable Data for Complex User Intents (2026.acl-long)

Copied to clipboard

Challenge: Tool-calling agents are increasingly deployed in real-world customer-facing workflows . but most studies on tool-callers focus on idealized settings with general, fixed, and well-specified tasks.
Approach: They propose a tool-calling agent-based data pipeline that converts trajectories into user-facing tasks with controlled intent adaptations.
Outcome: The proposed pipeline can be used to study tool use under three scenarios.
A Bounding Box is Worth One Token - Interleaving Layout and Text in a Large Language Model for Document Understanding (2025.findings-acl)

Copied to clipboard

Challenge: Existing methods for integrating spatial layouts with text have limitations . existing methods produce overly long text sequences or lack autoregressive traits of LLMs .
Approach: They introduce Interleaving Layout and Text in a Large Language Model (LayTextLLM) they use OCR-derived text and spatial layouts to integrate with LLMs for document understanding .
Outcome: The proposed model shows an increase in performance in KIE and VQA tasks.
SMARTMiner: Extracting and Evaluating SMART Goals from Low-Resource Health Coaching Notes (2025.findings-emnlp)

Copied to clipboard

Challenge: SMARTMiner extracts specific, measurable, attainable, relevant, time-bound (SMART) goals from unstructured health coaching notes.
Approach: They propose a framework for extracting and evaluating specific, measurable, attainable, relevant, time-bound (SMART) goals from unstructured health coaching notes.
Outcome: The framework extracts behavior change goal spans and categorizes their SMARTness.
Can LLMs Hear the Dogwhistle? (2026.findings-acl)

Copied to clipboard

Challenge: Existing safety benchmarks focus on explicitly harmful content, but ignore context-dependent expressions such as dogwhistles.
Approach: They propose a benchmark for evaluating LLM safety under dogwhistle-driven prompts . their findings expose a blind spot in current safety evaluation practices .
Outcome: The proposed benchmark compared safety performance with toxic terms using dogwhistle-driven prompts.
MMSearch-R1: Incentivizing LMMs to Search (2026.acl-long)

Copied to clipboard

Challenge: Existing approaches to deploying large multimodal models rely on rigid pipelines . Existing methods such as retrieval-augmented generation and prompt engineered search rely only on rigid knowledge sources.
Approach: They propose a framework that enables LMMs to perform on-demand, multi-turn search in real-world Internet environments.
Outcome: The proposed model outperforms existing models while reducing search calls by over 30%.
LMMs-Eval: Reality Check on the Evaluation of Large Multimodal Models (2025.findings-naacl)

Copied to clipboard

Challenge: Current large foundational models have demonstrated transformative capabilities, approaching or surpassing human-level performances in many tasks.
Approach: They propose a unified and standardized multimodal benchmark framework with over 50 tasks and more than 10 models to promote transparent and reproducible evaluations.
Outcome: The proposed framework has 50 tasks and more than 10 models to promote transparent and reproducible evaluations.
Premise-based Multimodal Reasoning: Conditional Inference on Joint Textual and Visual Clues (2022.acl-long)

Copied to clipboard

Challenge: Existing work in vision language cross-modal reasoning uses binary or multi-choice classification based on source image and textual query.
Approach: They propose a task where a textual premise is the background presumption on each source image.
Outcome: The proposed task is based on a dataset of 15,360 movie screenshots and human-curated premise templates from 6 pre-defined categories.
Multimodal Emotion Recognition in Conversations: A Survey of Methods, Trends, Challenges and Prospects (2025.findings-emnlp)

Copied to clipboard

Challenge: Multimodal Emotion Recognition in Conversations (MERC) is a new way to enhance human-computer interaction.
Approach: This survey offers a systematic overview of Multimodal Emotion Recognition in Conversations . it examines motivations, core tasks, representative methods, and evaluation strategies .
Outcome: The survey examines the effectiveness of MERC and its evaluation strategies.
MMInA: Benchmarking Multihop Multimodal Internet Agents (2025.findings-acl)

Copied to clipboard

Challenge: Existing benchmarks fail to assess embodied agents in a realistic, evolving environment for compositional Internet tasks.
Approach: They propose a multihop and multimodal benchmark to evaluate embodied agents for compositional Internet tasks.
Outcome: The proposed protocol significantly improves the performance of both the single-hop and multihop web browsing abilities.
Not All Voices Are Rewarded Equally: Probing and Repairing Reward Models across Human Diversity (2025.findings-emnlp)

Copied to clipboard

Challenge: Using real-world datasets, we conduct the most comprehensive study to date, auditing various state-of-the-art reward models across nine sensitive attributes, including age, gender, ethnicity, etc.
Approach: They propose a method to mitigate group disparities in reward modeling by using real-world data.
Outcome: The proposed method is based on a population-based dataset with nine demographic attributes, including gender, ethnicity, age, gender, and ethnicity.
Uni-MMMU: A Massive Multi-discipline Multimodal Unified Benchmark (2026.acl-long)

Copied to clipboard

Challenge: Existing evaluations treat visual understanding and generation in isolation or overlook tasks that inherently couple them.
Approach: They propose a benchmark that examines the bidirectional synergy between generation and understanding across eight reasoning-centric domains.
Outcome: The proposed model systematically unfolds the bidirectional synergy between generation and understanding across eight reasoning-centric domains.
MaXIFE: Multilingual and Cross-lingual Instruction Following Evaluation (2025.acl-long)

Copied to clipboard

Challenge: Existing evaluation methods focus on single-language scenarios, overlooking multilingual and cross-lingual contexts.
Approach: They propose a tool to assess instruction-following capabilities across 23 different languages with 1667 verifiable instruction tasks.
Outcome: MaXIFE evaluates instruction-following capabilities across 23 languages with 1667 verifiable instruction tasks.
Evaluation Agent: Efficient and Promptable Evaluation Framework for Visual Generative Models (2025.acl-long)

Copied to clipboard

Challenge: Existing evaluation methods rely on rigid pipelines that overlook user needs and provide numerical results without clear explanations.
Approach: They propose an evaluation framework that employs human-like strategies for efficient, dynamic, multi-round evaluations using only a few samples per round.
Outcome: The evaluation agent framework reduces evaluation time to 10% of traditional methods while delivering comparable results.
MGPO: Thinking with Images via Multi-Turn Grounding-Based Reinforcement Learning (2026.findings-acl)

Copied to clipboard

Challenge: State-of-the-art large multimodal models face challenges when processing high-resolution images, as these inputs are converted into enormous visual tokens, many of which are irrelevant to the downstream task.
Approach: They propose a multi-turn grounding-based policy optimization framework that enables LMMs to iteratively focus on key visual regions by automatically cropping sub-images based on model-predicted grounding coordinates within a multiple-turn conversation framework.
Outcome: The proposed framework improves on Qwen2.5-VL-7B with 21K samples and surpasses OpenAI’s o1 and GPT-4o models on the out-of-distribution (OOD) V* Bench.
Learn to Adapt for Generalized Zero-Shot Text Classification (2022.acl-long)

Copied to clipboard

Challenge: Existing methods for generalized zero-shot text classification generalize poorly since the learned parameters are only optimal for seen classes rather than for both classes.
Approach: They propose a network that trains an adaptive classifier by using both seen and virtual unseen classes to simulate a generalized zero-shot learning scenario.
Outcome: The proposed model outperforms several previous approaches on five text classification datasets.
A Survey on Open Information Extraction from Rule-based Model to Large Language Model (2024.findings-emnlp)

Copied to clipboard

Challenge: Open Information Extraction (OpenIE) is a key NLP task aimed at extracting structured information from unstructured text sources.
Approach: They propose to categorize OpenIE into rule-based, neural, and pre-trained large language models and discuss each within a chronological framework.
Outcome: The paper categorizes OpenIE approaches into rule-based, neural, and pre-trained large language models, discussing each within a chronological framework.
RHO: Reducing Hallucination in Open-domain Dialogues with Knowledge Grounding (2023.findings-acl)

Copied to clipboard

Challenge: Existing knowledge-grounded dialogue systems generate accurate and informative responses, but they are prone to hallucination problems.
Approach: They propose a method to generate hallucinated responses using knowledge graphs . they propose local knowledge grounding to combine textual embeddings with corresponding KG embeddments . a global knowledge ground technique is also proposed to equip RHO with multi-hop reasoning abilities .
Outcome: The proposed approach outperforms state-of-the-art methods on automatic and human evaluation by a large margin.
Plausible May Not Be Faithful: Probing Object Hallucination in Vision-Language Pre-training (2023.eacl-main)

Copied to clipboard

Challenge: Large-scale vision-language pre-trained (VLP) models generate unfaithful or nonsensical texts given the source input, which is called hallucination.
Approach: They propose a VLP loss-based model to mitigate object hallucination by decoupling VLP objectives and a token-level image-text alignment.
Outcome: The proposed model reduces object hallucination by 17.4% on two benchmarks.
A Review of Incorporating Psychological Theories in LLMs (2026.eacl-long)

Copied to clipboard

Challenge: a holistic review systematically integrating psychology across the LLM lifecycle remains missing.
Approach: They examine how psychological theories can inform stages of LLM development . they highlight current trends and gaps in how psychological theory is applied .
Outcome: The authors highlight current trends and gaps in how psychological theories are applied . they argue that psychological insights have shaped pivotal NLP breakthroughs .
Towards Scalable Lightweight GUI Agents via Multi-role Orchestration (2026.findings-acl)

Copied to clipboard

Challenge: Advanced GUI agents suffer from prohibitive deployment costs on resource-constrained devices.
Approach: They propose a lightweight GUI agent with GUI-specific knowledge and task scalability . LAMO-3B supports monolithic execution and MAS-style orchestration .
Outcome: The proposed GUI agent LAMO-3B supports monolithic execution and MAS-style orchestration.
CompTab: A Comprehensive Benchmark for Real-World TableQA with Complex Reasoning and Irregular Tables (2026.acl-long)

Copied to clipboard

Challenge: Existing benchmarks focus on well-structured tables and fail to reflect irregular structures and complex reasoning commonly encountered in real-world scenarios.
Approach: They propose a benchmark to evaluate TableQA under complex reasoning and irregular table conditions.
Outcome: The proposed framework improves generalization and realism of large language models under complex and irregular table conditions.
DongbaMIE: A Multimodal Information Extraction Dataset for Evaluating Semantic Understanding of Dongba Pictograms (2025.findings-emnlp)

Copied to clipboard

Challenge: Dongba pictographic is the only pictograph script still in use in the world.
Approach: DongbaMIE is the first dataset focusing on multimodal information extraction of Dongbe pictographs.
Outcome: The dataset contains 23,530 sentence-level and 2,539 paragraph-level high-quality text-image pairs.
Unsolvable Problem Detection: Robust Understanding Evaluation for Large Multimodal Models (2025.acl-long)

Copied to clipboard

Challenge: Multiple-choice question answering (MCQA) is widely used to assess the understanding capability of Large Multimodal Models (LMMs).
Approach: They propose a task to evaluate the robust understanding capability of Large Multimodal Models (LMMs) they introduce a benchmark to assess performance across various ability dimensions .
Outcome: The proposed model can withhold answers when encountering unsolvable problems of MCQA, proving it understands the answer.
VChain: Chain-of-Visual-Thought for Reasoning in Video Generation (2026.findings-acl)

Copied to clipboard

Challenge: Recent video generation models struggle to synthesize complex dynamics with a coherent chain of consequences.
Approach: They propose a framework that injects visual reasoning signals from multimodal models into video generation.
Outcome: a new framework that leverages multimodal models to generate sparse keyframes significantly improves quality of generated videos.

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