3 papers
cs.CL2026
A Step Towards Forgetting: Optimiser History and the Loss of Answer Mass
Valeria Ruscio, Seth Nabarro, Keiran Thompson
During fine-tuning, a language model can assign less probability to previously learned answers even when the current gradient acts to preserve that probability. With momentum, each…
cs.LG2026
The Geometry of Learning: Read-Write Asymmetry Across Training
Valeria Ruscio, Eli-Shaoul Khedouri, Keiran Thompson
Pretraining and alignment update the same transformer weights, and we ask in which directions each stage changes them. The residual stream gives every matrix a shared coordinate sy…
cs.AI2026
On the Tip of the Tongue: Why LLMs Hallucinate Answers They Can Decode
Valeria Ruscio, Keiran Thompson
A language model can give the wrong answer even when the correct answer is decodable from its intermediate states. To study this gap between decodability and selection, we distingu…