1 citations · 2 across the 14 of their papers we have counts for
15 papers
What Shapes Emergent Misalignment? Insights from Training Dynamics, Model Priors, and Data
Yuchen Zhang, Anietta Weckauff, Diego Garcia-Olano +1
Emergent misalignment (EM) is a phenomenon in which models generalize with narrow fine-tuning, leading to broad (yet uneven) misalignment across evaluation questions. We study EM a…
Characterizing the Consistency of the Emergent Misalignment Persona
Anietta Weckauff, Yuchen Zhang, Maksym Andriushchenko
Fine-tuning large language models (LLMs) on narrowly misaligned data generalizes to broadly misaligned behavior, a phenomenon termed emergent misalignment (EM). While prior work ha…
Language Family Matters: Evaluating LLM-Based ASR Across Linguistic Boundaries
Yuchen Zhang, Ravi Shekhar, Haralambos Mouratidis
Large Language Model (LLM)-powered Automatic Speech Recognition (ASR) systems achieve strong performance with limited resources by linking a frozen speech encoder to a pretrained L…
Loquetier: A Virtualized Multi-LoRA Framework for Unified LLM Fine-tuning and Serving
Yuchen Zhang, Hanyue Du, Chun Cao +1
Low-Rank Adaptation (LoRA) has become a widely adopted parameter-efficient fine-tuning (PEFT) technique for adapting large language models (LLMs) to downstream tasks. While prior w…
Adamas: Hadamard Sparse Attention for Efficient Long-Context Inference
Siyuan Yan, Guo-Qing Jiang, Yuchen Zhang +4
Large language models (LLMs) now support context windows of hundreds of thousands to millions of tokens, enabling applications such as long-document summarization, large-scale code…
Hallucination reduction with CASAL: Contrastive Activation Steering For Amortized Learning
Wannan, Yang, Xinchi Qiu +6
Large Language Models (LLMs) exhibit impressive capabilities but often hallucinate, confidently providing incorrect answers instead of admitting ignorance. Prior work has shown tha…
Toward Efficient Influence Function: Dropout as a Compression Tool
Yuchen Zhang, Mohammad Mohammadi Amiri
Assessing the impact the training data on machine learning models is crucial for understanding the behavior of the model, enhancing the transparency, and selecting training data. I…
Echo: Decoupling Inference and Training for Large-Scale RL Alignment on Heterogeneous Swarms
Jie Xiao, Changyuan Fan, Qingnan Ren +6
Modern RL-based post-training for large language models (LLMs) co-locate trajectory sampling and policy optimisation on the same GPU cluster, forcing the system to switch between i…
The Eloquence team submission for task 1 of MLC-SLM challenge
Lorenzo Concina, Jordi Luque, Alessio Brutti +2
In this paper, we present our studies and experiments carried out for the task 1 of the Challenge and Workshop on Multilingual Conversational Speech Language Model (MLC-SLM), which…
Adversarial Curriculum Graph-Free Knowledge Distillation for Graph Neural Networks
Yuang Jia, Xiaojuan Shan, Jun Xia +5
Data-free Knowledge Distillation (DFKD) is a method that constructs pseudo-samples using a generator without real data, and transfers knowledge from a teacher model to a student by…
HijackRAG: Hijacking Attacks against Retrieval-Augmented Large Language Models
Yucheng Zhang, Qinfeng Li, Tianyu Du +4
Retrieval-Augmented Generation (RAG) systems enhance large language models (LLMs) by integrating external knowledge, making them adaptable and cost-effective for various applicatio…
Enhancing Transferability of Adversarial Attacks with GE-AdvGAN+: A Comprehensive Framework for Gradient Editing
Zhibo Jin, Jiayu Zhang, Zhiyu Zhu +4
Transferable adversarial attacks pose significant threats to deep neural networks, particularly in black-box scenarios where internal model information is inaccessible. Studying ad…
A Benchmark Dataset for Multimodal Prediction of Enzymatic Function Coupling DNA Sequences and Natural Language
Yuchen Zhang, Ratish Kumar Chandrakant Jha, Soumya Bharadwaj +3
Predicting gene function from its DNA sequence is a fundamental challenge in biology. Many deep learning models have been proposed to embed DNA sequences and predict their enzymati…
Self-Control of LLM Behaviors by Compressing Suffix Gradient into Prefix Controller
Min Cai, Yuchen Zhang, Shichang Zhang +5
We propose SelfControl, an inference-time model control method utilizing gradients to control the behavior of large language models (LLMs) without explicit human annotations. Given…
Battle of the Large Language Models: Dolly vs LLaMA vs Vicuna vs Guanaco vs Bard vs ChatGPT -- A Text-to-SQL Parsing Comparison
Shuo Sun, Yuchen Zhang, Jiahuan Yan +4
The success of ChatGPT has ignited an AI race, with researchers striving to develop new large language models (LLMs) that can match or surpass the language understanding and genera…