5 papers
Stabilizing Efficient Reasoning with Step-Level Advantage Selection
Han Wang, Xiaodong Yu, Jialian Wu +4
Large language models (LLMs) achieve strong reasoning performance by allocating substantial computation at inference time, often generating long and verbose reasoning traces. While…
The Hot Mess of AI: How Does Misalignment Scale With Model Intelligence and Task Complexity?
Alexander Hägele, Aryo Pradipta Gema, Henry Sleight +2
As AI becomes more capable, we entrust it with more general and consequential tasks. The risks from failure grow more severe with increasing task scope. It is therefore important t…
Training Dynamics of the Cooldown Stage in Warmup-Stable-Decay Learning Rate Scheduler
Aleksandr Dremov, Alexander Hägele, Atli Kosson +1
Learning rate scheduling is essential in transformer training, where the final annealing plays a crucial role in getting the best performance. However, the mechanisms behind this c…
Inverse Scaling in Test-Time Compute
Aryo Pradipta Gema, Alexander Hägele, Runjin Chen +11
We construct evaluation tasks where extending the reasoning length of Large Reasoning Models (LRMs) deteriorates performance, exhibiting an inverse scaling relationship between tes…
The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training
Fabian Schaipp, Alexander Hägele, Adrien Taylor +2
We show that learning-rate schedules for large model training behave surprisingly similar to a performance bound from non-smooth convex optimization theory. We provide a bound for…