Papers by Yan Teng

11 papers
A Mousetrap: Fooling Large Reasoning Models for Jailbreak with Chain of Iterative Chaos (2025.findings-acl)

Copied to clipboard

Challenge: Large Reasoning Models (LRMs) have advanced beyond traditional Large Language Models, yet they pose heightened safety risks.
Approach: They propose a first jailbreak attack targeting Large Reasoning Models . they exploit a Chaos Machine component to transform attack prompts with diverse one-to-one mappings based on the reasoning chain .
Outcome: The proposed attack exploits the unique vulnerabilities of LRMs by integrating a Chaos Machine. success rates of the mousetrap attack are as high as 96%, 86% and 98% respectively.
Flames: Benchmarking Value Alignment of LLMs in Chinese (2024.naacl-long)

Copied to clipboard

Challenge: Existing benchmarks for large language models (LLMs) do not accurately uncover safety vulnerabilities in LLMs.
Approach: They propose a value alignment benchmark called Flames that encompasses both harmlessness principles and a unique morality dimension that integrates specific Chinese values such as harmony.
Outcome: The proposed model performs poorly on Flames, particularly in safety and fairness dimensions.
Lightweight, Dynamic Graph Convolutional Networks for AMR-to-Text Generation (2020.emnlp-main)

Copied to clipboard

Challenge: AMR-to-text generation is used to transduce Abstract Meaning Representation structures (AMRs) Graph Convolution Networks (GCNs) are not able to capture non-local information and follow a local (first-order) information aggregation scheme.
Approach: They propose a dynamic fusion mechanism that captures richer non-local interactions . they propose weight tied convolutions and group graph convolution to reduce memory usage .
Outcome: The proposed model outperforms state-of-the-art models on two benchmark datasets with significantly fewer parameters while maintaining the model capacity.
AgenticEval: Toward Agentic and Self-Evolving Safety Evaluation of Large Language Models (2026.findings-acl)

Copied to clipboard

Challenge: Existing static safety evaluation methods are ill-equipped to address dynamic nature of AI risks and evolving regulations, creating a critical safety gap.
Approach: They propose a new paradigm of agentic safety evaluation reframing evaluation as a continuous and self-evolving process rather than a one-time audit.
Outcome: The proposed framework shows a consistent decline in model safety as the evaluation hardens.
Fake Alignment: Are LLMs Really Aligned Well? (2024.naacl-long)

Copied to clipboard

Challenge: Existing studies on large language models have shown that they are poorly aligned in practice.
Approach: They propose a framework to evaluate safety in large language models . they propose two new metrics to quantify fake alignment and obtain corrected performance estimation.
Outcome: The proposed framework and two metrics show that some models with purported safety are poorly aligned in practice.
Non-Autoregressive Document-Level Machine Translation (2023.findings-emnlp)

Copied to clipboard

Challenge: Existing non-autoregressive translation models struggle with document context and handling discourse phenomena.
Approach: They propose a simple but effective design of sentence alignment between source and target to improve their performance on document-level machine translation.
Outcome: The proposed model achieves high acceleration on documents and sentence alignment significantly enhances their performance.
CoVerRL: Breaking the Consensus Trap in Label-Free Reasoning via Generator-Verifier Co-Evolution (2026.acl-long)

Copied to clipboard

Challenge: Label-free reinforcement learning enables large language models to improve reasoning capabilities . but as training maximizes self-consistency, output diversity collapses, authors say . authors propose a framework where a single model alternates between generator and verifier roles .
Approach: They propose a framework where a model alternates between generator and verifier roles, bootstrapping each other.
Outcome: Experiments show that CoVerRL outperforms label-free baselines on reasoning benchmarks . the framework can be used to improve reasoning abilities without ground-truth supervision .
Probing the Safety Robustness of LLMs in Latent Space (2026.acl-long)

Copied to clipboard

Challenge: Despite substantial progress in safety alignment techniques, aligned large language models can still produce unsafe responses under minor internal perturbations.
Approach: They introduce Activation Steering Attack (ASA) and leverage the Negative Log-Likelihood (NLL) as a diagnostic signal to probe the local sensitivity of safety behaviors in latent space.
Outcome: The proposed method is model-agnostic and supervision-free, enabling a general and reproducible diagnostic metric for analyzing safety robustness.
From Evasion to Concealment: Stealthy Knowledge Unlearning for LLMs (2025.findings-acl)

Copied to clipboard

Challenge: Existing approaches to unlearning often treat nonsensical responses or template-based refusals as the unlearning target, making the process even more vulnerable to attacks and jailbreaks.
Approach: They propose a method that uses inverted facts to remove the need for auxiliary models or retaining data while avoiding leakage.
Outcome: Evaluated on the ToFU Knowledge Unlearning dataset using Llama2-7B-Chat and Phi-1.5, MEOW outperforms baselines in forgetting quality while preserving model utility.
How Well Do Text Embedding Models Understand Syntax? (2023.findings-emnlp)

Copied to clipboard

Challenge: Existing text embedding models have not addressed syntactic understanding challenges, highlighting ineffectiveness and enhancing generalization ability.
Approach: They propose to examine the ability of text embedding models to generalize across syntactic contexts.
Outcome: The proposed models exhibit high similarity socres at this simple task.
ESC-Eval: Evaluating Emotion Support Conversations in Large Language Models (2024.emnlp-main)

Copied to clipboard

Challenge: Emotion Support Conversation (ESC) is a crucial application for reducing stress and providing emotional guidance.
Approach: They re-organize 2,801 role-playing cards to define roles of role-players . they train a specific role- playing model called ESC-Role which behaves more like a confused person than GPT-4 .
Outcome: The proposed model behaves more like a confused person than GPT-4, and the model performs better than GPLs.

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