4 papers
Languages in Whisper-Style Speech Encoders Align Both Phonetically and Semantically
Ryan Soh-Eun Shim, Domenico De Cristofaro, Chengzhi Martin Hu +2
Cross-lingual alignment in pretrained language models enables knowledge transfer across languages. Similar alignment has been reported in Whisper-style speech encoders, based on sp…
Surgical, Cheap, and Flexible: Mitigating False Refusal in Language Models via Single Vector Ablation
Xinpeng Wang, Chengzhi Hu, Paul Röttger +1
Training a language model to be both helpful and harmless requires careful calibration of refusal behaviours: Models should refuse to follow malicious instructions or give harmful…
Look at the Text: Instruction-Tuned Language Models are More Robust Multiple Choice Selectors than You Think
Xinpeng Wang, Chengzhi Hu, Bolei Ma +2
Multiple choice questions (MCQs) are commonly used to evaluate the capabilities of large language models (LLMs). One common way to evaluate the model response is to rank the candid…
"My Answer is C": First-Token Probabilities Do Not Match Text Answers in Instruction-Tuned Language Models
Xinpeng Wang, Bolei Ma, Chengzhi Hu +5
The open-ended nature of language generation makes the evaluation of autoregressive large language models (LLMs) challenging. One common evaluation approach uses multiple-choice qu…