collaborators

9 papers

cs.LG2026

Thinking at the Right Size: Amortized Distillation Across Post-Trained LLMs

Yan Zhou, Sara Kangaslahti, Jonathan Geuter +4

Practical deployment of large language models (LLMs) requires families of post-trained variants---instruction-tuned, reasoning-tuned, and chat-style models---each at multiple sizes…

cs.LG2026

Understanding Layer Patching in Model Size Interpolation

Sara Kangaslahti, Jonathan Geuter, Nihal V. Nayak +3

Zero-shot model size interpolation aims to create new models of intermediate target sizes by combining existing models without additional training. Recent work on boomerang distill…

cs.LG2026

Stop Training for the Worst: Progressive Unmasking Accelerates Masked Diffusion Training

Jaeyeon Kim, Jonathan Geuter, David Alvarez-Melis +2

Masked Diffusion Models (MDMs) have emerged as a promising approach for generative modeling in discrete spaces. By generating sequences in any order and allowing for parallel decod…

cs.LG2026

Guided Speculative Inference for Efficient Test-Time Alignment of LLMs

Jonathan Geuter, Youssef Mroueh, David Alvarez-Melis

We propose Guided Speculative Inference (GSI), a novel algorithm for efficient reward-guided decoding in large language models. GSI combines soft best-of- test-time scaling with…

cs.LG2026

Boomerang Distillation Enables Zero-Shot Model Size Interpolation

Sara Kangaslahti, Nihal V. Nayak, Jonathan Geuter +3

Large language models (LLMs) are typically deployed under diverse memory and compute constraints. Existing approaches build model families by training each size independently, whic…

cs.LG2025

RoBoN: Routed Online Best-of-n for Test-Time Scaling with Multiple LLMs

Jonathan Geuter, Gregor Kornhardt

Best-of- is a widely used test-time scaling approach for LLM inference. Yet despite evidence that LLMs exhibit complementary strengths across tasks, traditionally best-of- re…