4 papers
Probabilistic Calibration Is a Trainable Capability in Language Models
Davide Baldelli, Sruthi Kuriakose, Maryam Hashemzadeh +2
Language models are increasingly used in settings where outputs should follow an intended distribution, yet their generation probabilities are often poorly calibrated to it. In man…
LLMs Can't Play Hangman: On the Necessity of a Private Working Memory for Language Agents
Davide Baldelli, Ali Parviz, Amal Zouaq +1
As LLMs move from text completion toward autonomous agents, they remain constrained by the standard chat interface, which lacks private working memory. This raises a fundamental qu…
Combining Domain and Alignment Vectors to Achieve Better Knowledge-Safety Trade-offs in LLMs
Megh Thakkar, Quentin Fournier, Matthew Riemer +4
There is a growing interest in training domain-expert LLMs that excel in specific technical fields compared to their general-purpose instruction-tuned counterparts. However, these…
A Deep Dive into the Trade-Offs of Parameter-Efficient Preference Alignment Techniques
Megh Thakkar, Quentin Fournier, Matthew D Riemer +4
Large language models are first pre-trained on trillions of tokens and then instruction-tuned or aligned to specific preferences. While pre-training remains out of reach for most r…