6 papers
Post-Training Language Models for Crosslingual Consistency
Tianyu Liu, Jirui Qi, Mrinmaya Sachan +3
Language models often respond inconsistently to translation-equivalent prompts across languages, undermining the reliability of multilingual systems. To quantify this, we give an i…
Ensembling Language Models with Sequential Monte Carlo
Robin Shing Moon Chan, Tianyu Liu, Samuel Kiegeland +5
Practitioners have access to an abundance of language models and prompting strategies for solving many language modeling tasks; yet prior work shows that modeling performance is hi…
Transducing Language Models
Vésteinn Snæbjarnarson, Samuel Kiegeland, Tianyu Liu +3
Modern language models define distributions over strings, but downstream tasks often require different output formats. For instance, a model that generates byte-pair strings does n…
Pointwise Mutual Information as a Performance Gauge for Retrieval-Augmented Generation
Tianyu Liu, Jirui Qi, Paul He +3
Recent work suggests that large language models enhanced with retrieval-augmented generation are easily influenced by the order, in which the retrieved documents are presented to t…
A Probability--Quality Trade-off in Aligned Language Models and its Relation to Sampling Adaptors
Naaman Tan, Josef Valvoda, Tianyu Liu +4
The relationship between the quality of a string, as judged by a human reader, and its probability, under a language model undergirds the development of better…
Efficiently Computing Susceptibility to Context in Language Models
Tianyu Liu, Kevin Du, Mrinmaya Sachan +1
One strength of modern language models is their ability to incorporate information from a user-input context when answering queries. However, they are not equally sensitive to the…