6 papers · 1 filter
How Persuasive is Your Context?
Tu Nguyen, Kevin Du, Alexander Miserlis Hoyle +1
Two central capabilities of language models (LMs) are: (i) drawing on prior knowledge about entities, which allows them to answer queries such as "What's the official language of A…
A Latent-Variable Model for Intrinsic Probing
Karolina StaÅczak, Lucas Torroba Hennigen, Adina Williams +2
The success of pre-trained contextualized representations has prompted researchers to analyze them for the presence of linguistic information. Indeed, it is natural to assume that…
Controllable Context Sensitivity and the Knob Behind It
Julian Minder, Kevin Du, Niklas Stoehr +4
When making predictions, a language model must trade off how much it relies on its context vs. its prior knowledge. Choosing how sensitive the model is to its context is a fundamen…
A Geometric Notion of Causal Probing
Clément Guerner, Tianyu Liu, Anej Svete +2
The linear subspace hypothesis (Bolukbasi et al., 2016) states that, in a language model's representation space, all information about a concept such as verbal number is encoded in…
Activation Scaling for Steering and Interpreting Language Models
Niklas Stoehr, Kevin Du, Vésteinn Snæbjarnarson +3
Given the prompt "Rome is in", can we steer a language model to flip its prediction of an incorrect token "France" to a correct token "Italy" by only multiplying a few relevant act…
Context versus Prior Knowledge in Language Models
Kevin Du, Vésteinn Snæbjarnarson, Niklas Stoehr +3
To answer a question, language models often need to integrate prior knowledge learned during pretraining and new information presented in context. We hypothesize that models perfor…