Papers by Gongshen Liu

25 papers
EVA: Evolving Semantic Adversaries for Red-Teaming GUI Agents Against Environmental Injection Attacks (2026.findings-acl)

Copied to clipboard

Challenge: Existing methods for red-teaming face a trade-off between requiring target-specific knowledge and incurring prohibitive computational costs.
Approach: They propose a framework that evolves payloads exclusively on the semantic dimension via a discovery-deployment pipeline.
Outcome: Experiments show that EVA outperforms baselines in terms of attack success rate while evolving benign seeds into successful attacks within 1.18 to 1.71 iterations.
The Confidence Paradox: Unveiling the Latent Discriminative Power of Diffusion Large Language Models in Mathematical Reasoning (2026.findings-acl)

Copied to clipboard

Challenge: Diffusion large language models (DLLMs) are a promising alternative to autoregressive (AR) generation, offering token-level probabilities under bidirectional context.
Approach: They propose to use diffusion large language models to generate token-level probabilities under bidirectional context and to examine the calibration paradox inherent to their native uncertainty estimates.
Outcome: The proposed model outperforms AR baselines on mathematical reasoning benchmarks and is highly miscalibrated on reasoning benchmark.
UOR: Universal Backdoor Attacks on Pre-trained Language Models (2024.findings-acl)

Copied to clipboard

Challenge: Existing methods to attack pre-trained language models rely on manual selection of triggers and backdoor representations.
Approach: They propose a backdoor attack method that turns manual selection into automatic optimization . they propose to use poisoned contrastive learning to learn more uniform backdoor representations .
Outcome: The proposed method achieves better attack performance on text classification tasks compared to manual methods.
SEARL: Joint Optimization of Policy and Tool Graph Memory for Self-Evolving Agents (2026.acl-long)

Copied to clipboard

Challenge: Recent advances in Reinforcement Learning with Verifiable Rewards (RLVR) have demonstrated significant potential in single-turn reasoning tasks.
Approach: They propose a tool-memory based self-evolving agentic framework that integrates planning with execution.
Outcome: The proposed framework is able to extract explicit knowledge from historical data and leverage inter-trajectory correlations to densify reward signals.
Investigating Multi-Hop Factual Shortcuts in Knowledge Editing of Large Language Models (2024.acl-long)

Copied to clipboard

Challenge: Recent work has demonstrated the power of large language models in recalling knowledge and reasoning.
Approach: They propose to erase shortcut neurons to mitigate the associated risks . 20% of the failures are attributed to shortcuts, they find .
Outcome: The proposed approach reduces failures in multi-hop knowledge editing caused by shortcuts by 20% .
Backdoor NLP Models via AI-Generated Text (2024.lrec-main)

Copied to clipboard

Challenge: Existing attacks disregard fluency and semantic fidelity of poisoned text, rendering it easily detectable.
Approach: They propose to use AI-generated poisoned text to attack NLP models by establishing covert associations between trigger patterns and target labels without affecting normal accuracy.
Outcome: The proposed method achieves effective attacks while maintaining fluency and semantic similarity across all scenarios.
TransAdv: A Translation-based Adversarial Learning Framework for Zero-Resource Cross-Lingual Named Entity Recognition (2022.findings-emnlp)

Copied to clipboard

Challenge: Existing methods for named entity recognition are limited by noise in translation . Existing approaches to named entities recognition are mainly based on labeled data .
Approach: They propose a framework to mitigate lexical and syntactic errors of translated data . they propose to use multi-level adversarial learning and multi-model knowledge distillation to mitigate noise .
Outcome: The proposed framework mitigates lexical and syntactic errors of translated data . it achieves competitive performance to state-of-the-art models .
OS-Kairos: Adaptive Interaction for MLLM-Powered GUI Agents (2025.findings-acl)

Copied to clipboard

Challenge: Existing efforts to build GUI agents focused on the autonomous mode have failed to address the problem of over-execution.
Approach: They propose an adaptive GUI agent that predicts confidence levels at each interaction step and elicits adaptive interaction.
Outcome: The proposed GUI agent outperforms existing models on a complex dataset and on established benchmarks.
Hierarchy-Aware Global Model for Hierarchical Text Classification (2020.acl-main)

Copied to clipboard

Challenge: Existing methods for hierarchical text classification are limited and lack holistic structural information.
Approach: They propose a hierarchy-aware global model with two variants that learn hierarchy-based label embeddings through an encoder and conduct inductive fusion of label-alike text features.
Outcome: The proposed model improves on three benchmark datasets.
Gracefully Filtering Backdoor Samples for Generative Large Language Models without Retraining (2025.coling-main)

Copied to clipboard

Challenge: Existing backdoor defense methods are ineffective for generative large language models . generative LLMs output sequences of high-dimensional token logits instead of low-dimensional classification logits .
Approach: They propose a method that leverages sample-wise gradients to identify backdoor samples without retraining LLMs.
Outcome: The proposed method outperforms baselines significantly in identifying backdoor samples without retraining LLMs.
Entity-Aware Abstractive Multi-Document Summarization (2021.findings-acl)

Copied to clipboard

Challenge: Existing models for multidocument summarization do not focus on explicitly modeling the underlying semantic information across documents.
Approach: They propose an entityaware model for abstractive multi-document summarization that augments the classical Transformer-based encoder-decoder framework with a heterogeneous graph consisting of text units and entities as nodes.
Outcome: The proposed model can deal with saliency and redundancy issues explicitly and can be used with pre-trained language models, arriving at improved performance.
A Unified Syntax-aware Framework for Semantic Role Labeling (D18-1)

Copied to clipboard

Challenge: Syntactic information has been paid a great attention over the role of enhancing SRL . but the gap between syntax-aware and syntax-gnostic SRL is smaller . a new framework proposes syntax-based SRL for a wide range of NLP tasks .
Approach: They propose to extend existing models to investigate more effective ways of incorporating syntax into sequential neural networks.
Outcome: The proposed framework outperforms existing models on CoNLL-2009 benchmarks in English and Chinese.
SynGhost: Invisible and Universal Task-agnostic Backdoor Attack via Syntactic Transfer (2025.findings-naacl)

Copied to clipboard

Challenge: Existing attacks are classified into end-to-end and pre-training types based on the attack phase . Existing backdoor attacks are based upon perplexity, fine-pruning, and maxEntropy.
Approach: They propose an entropy-based poisoning filter that mitigates backdoor attacks . they propose an invisible and universal task-agnostic backdoor attack via syntactic transfer .
Outcome: The proposed attack can transfer backdoors to various downstream tasks while preserving pre-trained language models' pre-training capabilities.
Hidden Ghost Hand: Unveiling Backdoor Vulnerabilities in MLLM-Powered Mobile GUI Agents (2025.findings-emnlp)

Copied to clipboard

Challenge: MLLM-powered GUI agents expose multiple interaction-level triggers, causing backdoor attacks . backdoor injection maximizes feature difference across sample classes, improving flexibility .
Approach: They propose a framework for red-teaming backdoor attacks using MLLMs . they construct composite triggers by combining goal and interaction levels .
Outcome: The proposed framework is effective and stealthy for red-teaming backdoor attacks.
Multiple Character Embeddings for Chinese Word Segmentation (P19-2)

Copied to clipboard

Challenge: Chinese word segmentation is regarded as character-based sequence labeling task in most current work but it neglects important fact: Chinese characters contain both semantic and phonetic meanings.
Approach: They propose a shared bi-LSTM-CRF model which fuses linguistic features efficiently by sharing the LSTM network during the training procedure.
Outcome: The proposed model achieves state-of-the-art in AS and CityU corpora without external lexical resources.
A Multi-Task Dual-Tree Network for Aspect Sentiment Triplet Extraction (2022.coling-1)

Copied to clipboard

Challenge: Existing methods are poor at detecting complicated relations between aspects and opinions . detecting unclear boundaries of multi-word aspects and opinion is also a challenge .
Approach: They propose a multi-task dual-tree network to extract triplets from a given sentence . they employ a constituency tree and a modified dependency tree to enhance the interaction .
Outcome: The proposed model extracts triplets from a given sentence, and it is effective on four datasets.
Modeling Multi-turn Conversation with Deep Utterance Aggregation (C18-1)

Copied to clipboard

Challenge: Existing work on retrieval-based context modeling for multi-turn conversation ignores interactions among previous utterances.
Approach: They propose retrieval-based response matching for multi-turn conversation . they propose to combine previous utterances into context using a deep utterrance aggregation model .
Outcome: The proposed model outperforms state-of-the-art methods on three multi-turn conversation benchmarks including an e-commerce dialogue corpus.
Acquiring Clean Language Models from Backdoor Poisoned Datasets by Downscaling Frequency Space (2024.acl-long)

Copied to clipboard

Challenge: Prior work attempts to mitigate backdoor learning during training LMs on poisoned datasets . backdoor attack poisons a small portion of training data by implanting specific text patterns .
Approach: They propose a multi-scale low-rank adaptive model that prioritizes learning of clean mapping . they propose radial scalings to reduce the success rate of diverse backdoor attacks .
Outcome: The proposed model outperforms baselines significantly in the frequency space . it reduces the success rate of diverse backdoor attacks to below 15% across datasets .
ALIS: Aligned LLM Instruction Security Strategy for Unsafe Input Prompt (2025.coling-main)

Copied to clipboard

Challenge: Existing instruction tuning methods may fail to balance performance with robustness against attacks from user input like prompt injection and jailbreaking.
Approach: They propose an instruction tuning paradigm to decompose user inputs into irreducible atomic instructions and organize them into instruction streams to guide response generation of model.
Outcome: The proposed model can maintain security constraints by ignoring or rejecting user mode instructions when user mode instruction conflicts with kernel mode instructions.
How Large Language Models Encode Context Knowledge? A Layer-Wise Probing Study (2024.lrec-main)

Copied to clipboard

Challenge: Existing studies have focused on enhancing the factualness of large language models using context knowledge.
Approach: They propose to use ChatGPT to construct probing datasets that provide diverse and coherent evidence corresponding to various facts.
Outcome: The proposed model can encode knowledge across different layers, and it is compared with existing models.
R-Judge: Benchmarking Safety Risk Awareness for LLM Agents (2024.findings-emnlp)

Copied to clipboard

Challenge: Large language models (LLMs) have shown compelling abilities in reasoning, decision-making, and instruction following.
Approach: They propose a benchmark to evaluate the proficiency of large language models (LLMs) in judging and identifying safety risks given agent interaction records.
Outcome: The proposed model outperforms the best-performing model, GPT-4o, while no other models significantly exceed the random.
Is Continuous Prompt a Combination of Discrete Prompts? Towards a Novel View for Interpreting Continuous Prompts (2023.findings-acl)

Copied to clipboard

Challenge: Existing studies on the interpretability and transferability of continuous prompts have not been conducted on the subject.
Approach: They propose to interpret continuous prompts as the weighting of discrete prompts by jointly optimizing prompt fidelity and downstream fidelity.
Outcome: The proposed interpretations provide effective readability and plausibility, which is helpful to understand the decision-making of continuous prompts and discover potential shortcuts.
Few-shot Table-to-text Generation with Prefix-Controlled Generator (2022.coling-1)

Copied to clipboard

Challenge: Neural table-to-text generation approaches are data-hungry and lack labeled data.
Approach: They propose a prompt-based approach for few-shot table-to-text generation using a task-specific prefix and an input-specific input prefix.
Outcome: The proposed approach is able to generate table-to-text summaries with a few instances and is validated on human, book and song datasets.
Improving Constituent Representation with Hypertree Neural Networks (2022.naacl-main)

Copied to clipboard

Challenge: Existing methods of span representation are based on simple derivations from word representations and do not utilize compositional structures of natural language.
Approach: They propose a hypertree neural network that is structured with constituency parse trees to improve representations of constituent spans.
Outcome: The proposed model improves representations of constituent spans using constituency parse trees.
Sliced Recurrent Neural Networks (C18-1)

Copied to clipboard

Challenge: Recurrent neural networks have difficulty in parallelization because of their recurrent structure.
Approach: They propose sliced recurrent neural networks (SRNNs) which can be parallelized by slicing sequences into many subsequences.
Outcome: The proposed recurrent neural networks perform better than standard RNNs on six large-scale sentiment analysis 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