Publications (20)
MIR-Bench: Can Your LLM Recognize Complicated Patterns via Many-Shot In-Context Reasoning?
Kai Yan, Zhan Ling, Kang Liu +5
The ability to recognize patterns from examples and apply them to new ones is a primal ability for general intelligence, and is widely studied by psychology and AI researchers. Man…
DiffNorm: Self-Supervised Normalization for Non-autoregressive Speech-to-speech Translation
Weiting Tan, Jingyu Zhang, Lingfeng Shen +2
Non-autoregressive Transformers (NATs) are recently applied in direct speech-to-speech translation systems, which convert speech across different languages without intermediate tex…
LongReason: A Synthetic Long-Context Reasoning Benchmark via Context Expansion
Zhan Ling, Kang Liu, Kai Yan +6
Large language models (LLMs) have demonstrated remarkable progress in understanding long-context inputs. However, benchmarks for evaluating the long-context reasoning abilities of…
Reliable and Responsible Foundation Models: A Comprehensive Survey
Xinyu Yang, Junlin Han, Rishi Bommasani +49
Foundation models, including Large Language Models (LLMs), Multimodal Large Language Models (MLLMs), Image Generative Models (i.e, Text-to-Image Models and Image-Editing Models), a…
Sen2Pro: A Probabilistic Perspective to Sentence Embedding from Pre-trained Language Model
Lingfeng Shen, Haiyun Jiang, Lemao Liu +1
Sentence embedding is one of the most fundamental tasks in Natural Language Processing and plays an important role in various tasks. The recent breakthrough in sentence embedding i…
On the Evaluation Metrics for Paraphrase Generation
Lingfeng Shen, Lemao Liu, Haiyun Jiang +1
In this paper we revisit automatic metrics for paraphrase evaluation and obtain two findings that disobey conventional wisdom: (1) Reference-free metrics achieve better performance…
It Takes Two: On the Seamlessness between Reward and Policy Model in RLHF
Taiming Lu, Lingfeng Shen, Xinyu Yang +3
Reinforcement Learning from Human Feedback (RLHF) involves training policy models (PMs) and reward models (RMs) to align language models with human preferences. Instead of focusing…
Flatness-Aware Prompt Selection Improves Accuracy and Sample Efficiency
Lingfeng Shen, Weiting Tan, Boyuan Zheng +1
With growing capabilities of large language models, prompting them has become the dominant way to access them. This has motivated the development of strategies for automatically se…
Do pretrained Transformers Learn In-Context by Gradient Descent?
Lingfeng Shen, Aayush Mishra, Daniel Khashabi
The emergence of In-Context Learning (ICL) in LLMs remains a remarkable phenomenon that is partially understood. To explain ICL, recent studies have created theoretical connections…
The Language Barrier: Dissecting Safety Challenges of LLMs in Multilingual Contexts
Lingfeng Shen, Weiting Tan, Sihao Chen +6
As the influence of large language models (LLMs) spans across global communities, their safety challenges in multilingual settings become paramount for alignment research. This pap…
TextShield: Beyond Successfully Detecting Adversarial Sentences in Text Classification
Lingfeng Shen, Ze Zhang, Haiyun Jiang +1
Adversarial attack serves as a major challenge for neural network models in NLP, which precludes the model's deployment in safety-critical applications. A recent line of work, dete…
AnaloBench: Benchmarking the Identification of Abstract and Long-context Analogies
Xiao Ye, Andrew Wang, Jacob Choi +6
Humans regularly engage in analogical thinking, relating personal experiences to current situations (X is analogous to Y because of Z). Analogical thinking allows humans to solve p…
Generalizable End-to-End Tool-Use RL with Synthetic CodeGym
Weihua Du, Hailei Gong, Zhan Ling +7
Tool-augmented large language models (LLMs), hereafter LLM agents, leverage external tools to solve diverse tasks and interface with the real world. However, current training pract…
Narrowing the Gap between Zero- and Few-shot Machine Translation by Matching Styles
Weiting Tan, Haoran Xu, Lingfeng Shen +5
Large language models trained primarily in a monolingual setting have demonstrated their ability to generalize to machine translation using zero- and few-shot examples with in-cont…
Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translation
Haoran Xu, Amr Sharaf, Yunmo Chen +5
Moderate-sized large language models (LLMs) -- those with 7B or 13B parameters -- exhibit promising machine translation (MT) performance. However, even the top-performing 13B LLM-b…
SemStamp: A Semantic Watermark with Paraphrastic Robustness for Text Generation
Abe Bohan Hou, Jingyu Zhang, Tianxing He +7
Existing watermarking algorithms are vulnerable to paraphrase attacks because of their token-level design. To address this issue, we propose SemStamp, a robust sentence-level seman…
Rethink the Evaluation for Attack Strength of Backdoor Attacks in Natural Language Processing
Lingfeng Shen, Haiyun Jiang, Lemao Liu +1
It has been shown that natural language processing (NLP) models are vulnerable to a kind of security threat called the Backdoor Attack, which utilizes a `backdoor trigger' paradigm…
Frequency-aware Dimension Selection for Static Word Embedding by Mixed Product Distance
Lingfeng Shen, Haiyun Jiang, Lemao Liu +1
Static word embedding is still useful, particularly for context-unavailable tasks, because in the case of no context available, pre-trained language models often perform worse than…
The Trickle-down Impact of Reward (In-)consistency on RLHF
Lingfeng Shen, Sihao Chen, Linfeng Song +5
Standard practice within Reinforcement Learning from Human Feedback (RLHF) involves optimizing against a Reward Model (RM), which itself is trained to reflect human preferences for…
A Simple and Plug-and-play Method for Unsupervised Sentence Representation Enhancement
Lingfeng Shen, Haiyun Jiang, Lemao Liu +1
Generating proper embedding of sentences through an unsupervised way is beneficial to semantic matching and retrieval problems in real-world scenarios. This paper presents Represen…